Executive Summary... 3 Research Design... 3 Summary of Results... 3 Introduction... 6 Purpose... 6 Objectives... 6 Colombia Background Information...

C2/20%,$

Executive Summary... 3 Research Design... 3 Summary of Results... 3 Introduction... 6 Purpose... 6 Objectives... 6 Colombia Background Information... 7 Project Components... 8 Definitions... 8 Personnel... 9 Methods... 10 Survey... 10 Sample Size... 10 Sampling Strategy... 11 Data Collection... 12 Results... 15 Describing the Sample... 15 Ministry Focus... 24 Programs (Churches and NGOs)... 28 Program Categories... 39 Technology... 48 Secondary Impact... 49 Reach... 52 Limitations... 59 Acknowledgements... 60 Appendix A... 62 OneHope s Definition of Missional Reach... 62 Appendix B... 63 Colombia Asset Mapping Survey... 63 Appendix C... 72 Key Ministries... 72 2

EXECUTIVE SUMMARY Research Design In April 2013, this national, market evaluation study was begun in Colombia to describe, measure, and evaluate Christian ministry activities and services directed toward children and youth between the ages of 5- and 19-years-old in the country. This study is part of a broader effort led by OneHope, involving 40 countries from 2012 2014. This study reveals a national picture of the children and youth ministry sector, highlighting (a) key stakeholders and exemplars, (b) priority and commitment placed on children and youth, (c) programs and activities utilized, and (d) the average number of children and youth being reached by the ministry sector. Nationally representative samples of Protestant churches and Christian non-government organizations (NGOs) were surveyed to allow for an approximate calculation of annual Reach the total number of children and youth reached with the Gospel 1 and to identify geographic or demographic gaps where children and youth are not being reached. This study also serves to describe the types of programs (including evangelism, mercy ministry, etc.) being implemented by churches and NGOs and their frequency of use. This national, market evaluation study is intended to provide information to leaders of ministries and churches so that they may efficiently direct their efforts and improve coordination and partnership to serve as many children and youth with the Gospel as resources allow. Results will also guide future OneHope efforts in Colombia. Summary of Results Describing the Sample Complete surveys were received from 388 ministry organizations (including 373 churches and 15 NGOs). This study included a diverse group of churches, representing a wide range of denominations, geographic locations, and church sizes. The number of churches surveyed in each region was proportional to the estimated Christian population size in each region. The majority were located in urban areas. Ministry Focus There were significant relationships uncovered between ministry type and age group focus of the ministries: For local churches, focus appears to decrease as age increases, while for NGOs, focus appears to remain relatively constant. This same trend was seen across all denominational affiliations. There were significant relationships between geographic regions and age focus groups for churches. For example, significantly greater proportions of churches in the Caribbean and Amazon regions reported focus on ages 5 11 and ages 12 15 than did churches in the Pacific region. 1 One limitation of the analysis is the potential overlap in Reach among NGOs and churches, who may be reaching the same children and youth. 3

There were significant relationships between recoded church attendance and age focus groups: Larger churches reported a greater focus on children ages 12-15 and youth ages 12 19 than medium or small churches. Programs The most frequently offered programs by NGOs were programs focused on ministry to at-risk children (ages 5 11) and programs focused on leadership training/development for youth (ages 12 19). The most frequently offered programs by churches were Sunday school programs. This was true for both age groups. Churches located in the Andean region were significantly more likely to provide Bible clubs or studies, person-to-person evangelism, and leadership training/development programs for children ages 5 11 than were churches in other regions. Churches in the Amazon region were the least likely among all churches to hold large evangelistic gatherings/festivals for either children or youth. Churches in the Pacific region were significantly more likely than churches in any other region to conduct literacy programs for youth ages 12 19. Churches in areas with higher population densities (urban and suburban) were more likely than churches in more sparsely populated areas (rural) to offer leadership training/development programs for children ages 5 11. Surprisingly, recovery from addiction or trauma programs for youth ages 12 19 were offered by churches in rural areas at a significantly higher rate than were offered by churches in urban areas; however, sample sizes of rural and suburban churches were small. Program Categories Each program for children and youth was assigned to one of five categories: evangelism, discipleship, benevolence, evangelism and/or discipleship, or evangelism and/or benevolence. Almost all churches conducted evangelism and discipleship programs; however, they were less likely to conduct benevolence programs for children and youth. There was a significant relationship between program category and recoded denominational affiliation of churches offering programs for youth ages 12 19: Other Evangelical churches reported conducting more evangelism/benevolence programs than Independent, Inter- Denominational, Non-Denominational, and Baptist/CMA churches. Significant relationships were found between market population and discipleship programs for ages 5-11: Churches in urban and suburban areas were more likely to conduct discipleship programs for children ages 5-11 than were churches in rural areas. There were statistically significant relationships recoded church attendance and: evangelism programs for ages 5 11, evangelism programs for ages 12 19, discipleship programs for ages 12 19, benevolence programs for ages 12 19, evangelism/discipleship programs for ages 12 19, and evangelism/benevolence programs for ages 12 19. All program category proportions were higher for churches with more than 200 attendees than were the proportions for churches with fewer attendees. Technology Churches were most likely to use video and social media but least likely to use websites. Large proportions of youth ages 12 19 were reported to have cell phones, easy internet access, and video equipment. 4

Secondary Impact Ministries were asked about programs or activities that do not necessarily target, but may still impact, children and youth, such as life skills training, health programs, food programs, and water or sanitation programs. Churches and NGOs were most likely to conduct food programs. One of the four programs was significantly different across market population sizes: Churches in urban areas were less likely to conduct life skills than were churches in suburban areas. Also, churches in the Amazon region were significantly more likely to conduct life skills training than the churches in the Caribbean and Andean regions. Reach Based on self-reported Reach numbers across the national sample, Reach was estimated for children and youth across Colombian Protestant churches. Direct Gospel Reach (not including indirect Reach, such as Reach by radio, etc.) for the NGO sample was 46,630. This is a self-reported cumulative figure, which was based on targeting the largest Christian, child and youth focused NGOs in Colombia. Direct Gospel Reach for churches was calculated by multiplying median church Reach (18 for children and youth combined) by the number of Protestant and Evangelical churches reported by Operation World (20,043). Adding the total national church Reach (360,774) to the total national NGO Reach (46,630) estimates a cumulative Reach of 407,404 2. This number is limited by the inability to examine broader trends in media, possibly under-surveying key NGOs, and the decision to limit church surveys to Protestant congregations. 2 Reach calculations should be considered as preliminary and are subject to further adjustment based on such factors as incorporating actual mega-church reach and further clarification from participating ministries. 5

INTRODUCTION Purpose OneHope s mission is to affect destiny by providing God s eternal Word to all the children and youth of the world. That can be accomplished only through collaboration with the many churches and other organizations that are ministering to children and youth. As a tool to catalyze that kind of collaboration, OneHope is sponsoring this global study to describe the current state of institutional (church and NGO) ministry to children and youth (ages 5 19) in 40 countries. 3 The purpose of this study is twofold: Promote Understanding Complete a global study of the work of churches and Christian ministry organizations with children and youth to better understand which and how many children and youth are being reached with the Gospel. Catalyze Impact Identify key institutions serving children and youth as an informational and collaborative tool for pastors and practitioners toward better ministry, partnership, and impact. The three objectives of this study are as follows: Objectives To identify organizations that are involved, influential, important, impactful, and innovative, to share information, and to highlight people, programs, and practices that might serve the broader global church. 4 To describe the types of programs (e.g., evangelism, discipleship, and benevolence) being implemented to improve coordination among ministries and to identify both strengths and gaps in current ministry types and models, which may present opportunities for future ministry. To measure and map the percentage of children and youth being actively reached with the Gospel (Reach) in order to estimate progress in reaching all children and youth and to identify geographic or demographic gaps where children and youth are not being reached. To these ends, data were collected and will be popularized and made freely available to a broad national and international audience of ministry practitioners. 3 Partnership and feedback on the global project has been given by a number of key ministry networks and partner organizations invested in working with children and youth. 4 Except where authorization has been given to publish specific ministry names and/or data, confidentiality of all participating organizations is guaranteed. 6

Colombia Background Information Colombia 5 is located in northern South America. It borders both the Pacific Ocean and the Caribbean Sea, one on either side of its border with Panama. Venezuela and Brazil form Colombia s eastern borders, and Ecuador and Peru form the southern. The climate is mostly tropical along the coast and eastern plains regions and cooler toward the central highlands and Andes Mountains. Colombia has endured a long history of violence involving government and paramilitary groups. Within the past decade, a large number of paramilitary groups have dissolved, and the government is beginning to regain control of the nation. Spanish is the official language, and ethnic groups include Mestizo (58.0%), White (20.0%), Mulatto (14.0%), Black (4.0%), Mixed Black Amerindian (3.0%), and Amerindian (1.0%). Colombia s population of 45,745,783 is comprised of 92.5% Christian, 0.1% Muslim, 6.6% unaffiliated, and 0.8% folk religions (See Figure 0.1). Nearly 25.8% of Colombia s total population is comprised of youth ages 14 and under (See Figure 0.2), and according to a 2011 study, approximately 75.0% of the total population lives in an urban area, particularly Bogota, Medellin, Cali, Barranquilla, and Bucarmanga. Figure 0.1 Colombia Religions Muslim 0.1% Unaffiliated 6.6% Folk Religions 0.8% Christian 92.5% 5 This report relies upon population, demographic, and economic data from the United States Central Intelligence Agency World Factbook. (2014). Colombia. Retrieved from https://www.cia.gov/library/publications/the-worldfactbook/geos/co.html 7

Figure 0.2 Colombia Populations by Age 55 64 8.0% 65 Years & Over 6.5% 0 14 25.8% 25 54 41.5% 15 24 18.2% Administrative divisions include 32 departments and the capital district Bogota. Project Components Survey of Churches and Survey of NGOs A survey was designed and administered to pastors and leaders of churches, as well as to leaders and representatives of Christian ministry organizations reaching children and youth (NGOs) throughout Colombia. Interviews with Ministry Leaders In-depth interviews are being conducted with influential leaders of ministry organizations and churches during the 2014 calendar year. At the conclusion of these interviews, an updated report will be released. Definitions Ministries Christian organizations, including churches and NGOs that deliver programming (educational, religious, practical, or otherwise) to children and youth. NGOs Non government organizations, including non-profit organizations, schools, networks, foundations, and para-church organizations, etc. Reach Presenting the Gospel of salvation through Jesus Christ in the context of a direct, personal interaction where the child or youth has the opportunity to respond, ask questions, or pray with the presenter or a ministry representative. 6 6 Adapted from OneHope s Definition of Missional Reach (Appendix A). 8

Personnel The National Project Coordinator (NPC) facilitated this study and oversaw survey administration and interviews throughout Colombia. Survey administrators surveyed church and NGO representatives. Data entry staff transmitted survey data to the International Project Coordinator. The NPC or other trained interviewers interviewed church and NGO ministry leaders. 9

METHODS Survey The Colombia Asset Mapping Survey (Appendix B) was adapted from OneHope pilot studies performed in Uganda (2011) and the Philippines (2012), as well as similar studies conducted with the European Evangelical Alliance in four European Countries (2012). The survey was designed to be completed in paper form by any church, ministry, school, or other institution serving children or youth. On the survey, each organization was asked to provide the following information: Personal and ministry contact information City, town, or village demographics Religious affiliation Number of people employed by ministry Number of people regularly attending service (for churches only) Target age groups Number of children and youth impacted annually Ministry programs Ministry format (face-to-face vs. indirect; number of ministry opportunities with each child or youth; duration of ministry with each child or youth) Technology use by ministries and by served children and youth Geographic location(s) of ministry activities Usage of OneHope products Questions about the spiritual content of ministry activities were based on elements of OneHope s internal definition of missional Reach (Appendix A), which could be reasonably assessed through a survey. Sample Size Sampling methods differed for churches and NGOs. When surveying churches, the goal was to draw a representative sample of all churches in the country. The number of churches surveyed was sufficient to draw a representative sample of the estimated total number of churches 7 with a 95.0% level of confidence and a 5.0% margin of error. For example, in a country like Colombia with approximately 20,043 churches, a random sample of 377 churches was required. 7 As estimated by Operation World. Mandryk, & Jason. (2010). Operation World: The Definitive Prayer Guide to Every Nation (7th ed.). Westmont, IL: InterVarsity Press. 10

In contrast, due primarily to the inability to obtain a representative sample of NGO ministries, the project team eventually concluded that country level asset mapping is best served by inclusion of data from the largest NGOs willing to participate. In that way, findings from even a few organizations will be descriptive of the largest possible numbers of children and youth affected by the work of these organizations. Sampling Strategy A comprehensive national list of churches from which to draw a random sample was not available. Therefore, a proportional quota sample 8 was used in which the sample was representative of churches throughout the country. The desired number (or quota) of churches to survey was established for each geographic region, as described below. To increase representativeness of the sample, attempts were also made to include churches from a variety of denominations, urban and rural locations, and congregation sizes (including relatively small, medium, and large churches). Geographic Region States (or provinces, departments, etc.) were grouped into regions based on existing regional boundaries (geographic, religious, cultural, linguistic, or other similarities among states) and the advice of in-country experts. In most countries, sampling was conducted in all regions. Next, up to four states in each region were selected semi-randomly. The selection included states with a variety of population sizes within each region, but was biased toward the largest states. Finally, up to four counties (or local government areas, municipalities, districts, etc.) were selected semi-randomly within each selected state, following the same approach as that for selecting states. There is likely to be more Christian ministry to children and youth in places with a higher proportion of Christians. Therefore, the target sample size in each region, state, or county was proportional to the population size in that area, adjusted for the estimated percent of the population who are Christian 9 (more Christians equals a larger sample size, and vice versa) where data on religious affiliation were available. In most countries, data on religious affiliation were available for the regional level only. Denomination This study was focused on Protestant denominations, including Mainline Protestants, Evangelicals, Pentecostals, and Independents, as well as Orthodox organizations. When surveying churches and NGOs, survey administrators were encouraged to include a variety of denominations. The NPC was also provided with proposed denominational quotas for the survey of churches, which included the largest Protestant and Independent denominations in the country. 10 Denominational quotas were secondary to geographic quotas. Denominational quotas for the survey of churches were established in two steps. First, the quota for each denominational group (e.g., Mainline Protestant) was calculated to be proportional to 8 Food and Agriculture Organization. Examples of Sampling Methods. Downloaded 30 May, 2013 from http://www.fao.org/ag/humannutrition/32428-0613f516cb07eade922c8c19b4d0452c0.pdf 9 Based on estimates from the World Christian Database (downloaded May 2013 from http://www.worldchristiandatabase.org/wcd/) 10 Mandryk, & Jason. (2010). Operation World: The Definitive Prayer Guide to Every Nation (7th ed.). Westmont, IL: InterVarsity Press. 11

the estimated number of Christians in the country who were affiliated with that denominational group. 11 Second, quotas were calculated for the largest denominations (e.g., Baptist) within each group, in proportion to the estimated number of Christians affiliated with each denomination. 12 Because regional data on denominational affiliation were generally not available, only a national quota was used. Surveys were only administered to churches or NGOs that identified themselves as Christian. Unaffiliated Christian churches were classified as Protestant. Rural or Urban Location The number of churches to survey in rural and urban areas was proportional to the percentages of the national population living in rural and urban areas, respectively. Suburban churches were classified as urban. Data Collection The survey was administered to church pastors, youth pastors, ministry leaders, or other representatives of churches and Christian NGOs from August, 2013 to December, 2013. Identifying Churches to Survey In Colombia, in order to fulfill the church sampling quotas described above, survey administrators visited the selected states and counties within each region. However, flexibility was allowed in the location and number of surveys conducted at the state and county levels. This ensured an adequately diverse sample that was representative at both the regional and national level (in each report, results are presented only at the regional and national levels, not at the state or county levels). Survey administrators then identified churches to survey from among the eligible churches within each geographic area. Staff sought to include a variety of denominations and church sizes, as well as both rural and urban churches, and avoided being biased by personal preferences for certain denominations, worship styles, languages, etc. In other countries, the following sampling methods were also used as needed, while still adhering to the pre-determined quotas. Convenience Sampling The survey was distributed during relevant conferences or other gatherings, via network or denominational lists, or to pre-existing contacts. Random Selection When a listing of all churches in a country was available, churches were randomly selected and the survey was conducted by phone or in person. Identifying NGOs to Survey In addition to surveying the Christian NGOs reaching the largest number of children and youth in each country, in some cases the survey was also administered to NGOs identified via convenience sampling of contacts or networks, snowball sampling (in which respondents provide contact information for additional respondents to survey), or both. 11 Mandryk, & Jason. (2010). Operation World: The Definitive Prayer Guide to Every Nation (7th ed.). Westmont, IL: InterVarsity Press. 12 Ibid. 12

Data Entry and Evaluating Quotas Completed paper surveys were copied into an online database (using Survey Gizmo 13 ) as quickly as possible in order to allow ongoing monitoring of geographic quotas throughout the data collection process. When possible, geographic quotas were evaluated on a regular basis and when necessary, field sampling methods were adjusted to compensate for discrepancies in meeting quotas. The sample proportions sought to be within 5.0% of each of the quota established for each geographic region. In-Depth Interviews In-depth interviews will be conducted with 10 20 denominational leaders, national ministry leaders, and other influential individuals in service and ministry to Colombia s children and youth. Interviewees were identified by stakeholders and from recommendations received in completed surveys. When the analysis of interview data is complete, the results will provide insight into key areas of child and youth ministry in the country, including strengths and opportunities, weaknesses and challenges, recent trends, and examples of outstanding ministries. Analysis Data were entered from paper surveys into Survey Gizmo, an online survey development and data collection website. Exported data were analyzed in IBM SPSS (Statistical Package for the Social Sciences) and in Microsoft Excel. Data Cleaning Data on the number of children and youth reached annually were cleaned by averaging ranges (i.e., 200 300 becomes 250), converting non-numeric text to numeric responses ( two hundred becomes 200 ). For responses reporting over 1,000 children and youth reached, an investigation was conducted to guard against incorrect responses and possible misunderstandings. Calculating Reach Each organization was asked how many children and youth their ministry engages or interacts with. The numbers of children and youth reached, hereafter referred to as Reach, was adjusted for Content, Indirect Reach, and Time: Content Respondents were asked to estimate the number of children and youth that they reach who are presented the Gospel of salvation through Jesus Christ. Direct Reach Respondents were asked to estimate the number of children and youth that they have direct personal interaction with where the child or youth can respond to, ask questions, or pray with. Time Where an organization spent more than one year with a child or youth, Reach was reduced (e.g., if a ministry worked with each child for 3 years, annual Reach was divided by three). 13 http://www.surveygizmo.com/ 13

NGO Reach through Local Churches Where an NGO reported that all or some of its work was conducted through local churches, the total reach was reduced accordingly. This is based on an assumption that the children reached by local churches would be captured by the local church reach figures. After making the adjustments above, the national children and youth Reach of all churches in Colombia ( Church Reach ) was estimated by multiplying the median Reach for all churches in the survey by the estimated number of churches in Colombia. 14 National children and youth Reach was estimated by adding the Reach of surveyed NGOs to the national Church Reach. 14 As estimated by Operation World. Mandryk, & Jason. (2010). Operation World: The Definitive Prayer Guide to Every Nation (7th ed.). Westmont, IL: InterVarsity Press. 14

RESULTS Describing the Sample This section describes ministries that participated in the survey and includes statistical testing results to understand and describe relationships. The number of ministries that responded to specific questions is included in the charts/tables. Where statistically significant differences were identified, figures will show an asterisk (*) where appropriate, and details are provided in footnotes. A statistically significant finding is one that is determined (statistically) to be very unlikely to happen by chance. Statisticians are able to calculate the likelihood that any observed relationship between two variables (as indicated by any number of cases) could have happened by chance (or random variation). If it is calculated that there is less than a 1 in 20 chance (p.05) that the observed relationship could have happened by chance, the findings are considered significant. If there is less than a 1 in 100 chance (p.01), the findings are considered highly significant. Furthermore, a Cramer s V indicates the effect size or strength of the relationship between the variables. Cramer s V.251 à Very strong relationship.151.250 à Strong relationship.101.150 à Moderate relationship.051.100 à Weak relationship.010.050 à No or negligible relationship Frequency: Description of Churches and NGOs Frequency by Ministry Type As shown in Figure 1.1 below, of the ministries surveyed, 373 (96.1%) were local churches and 15 (3.9%) were NGOs. Figure 1.1 Ministry Type Local Churches 96.1% NGOs 3.9% N = 388 15

The study included a diverse group of churches, representing a wide range of denominations, geographic regions, and church sizes, as shown in the charts below. The majority of churches surveyed were located in urban areas. Participating NGOs were located in the urban areas of the Andean region, with one exception. Geography: Description of Churches and NGOs The geography of Colombia is characterized by its six main natural regions, as shown in Figure 1.2. 15 By design, the number of churches surveyed in each region was proportional to the estimated Christian population size in each region (for details, see the Methods section). In the survey data, 3 of the 6 regions contained fewer than 20 churches. This would not be a sufficient number of observations to perform analysis upon which to gain meaningful insights. Therefore, the one church from the island region of Insular was combined with the nearby coastal Caribbean region. In addition, the plains and rural areas of Orinoquia were combined with the Amazon jungle region contiguous to form a single Amazon region. Relative frequencies of observations, as well as spatial proximity and socio-cultural similarity were the primary considerations in aggregation decisions. The focus of this study was on surveying those NGOs which reach the largest number of children and youth. The vast majority of these large NGOs were headquartered in the Andean region. Therefore, as shown in Figure 1.3A, the Andean region has the highest NGO representation (93.3%) of surveyed ministries. 15 http://www.worldofmaps.net/en/south-america/map-colombia/map-region-colombia.htm 16

Geographic Regions Figure 1.2 Geographic Regions 17

Churches and NGOs by Geographic Region The numbers of churches and NGOs surveyed in each of the six geographic regions are summarized in Figure 1.3. Figure 1.3 Sample by Geographic Region Region Local Churches NGOs N % N % Orinoquia 14 3.8 0.0 Pacific 75 20.1 1 6.7 Caribbean 96 25.7 0.0 Andean 179 48.0 14 93.3 Amazon 8 2.1 0.0 Insular 1.3 0.0 Total 373 100.0 15 100.0 Churches and NGOs by Recoded Geographic Region As described above, ministry region was recoded from six to four categories to examine group differences. The numbers of churches and NGOs surveyed by recoded geographic region are summarized in Figure 1.4. Figure 1.4 Sample by Recoded Geographic Region Region Local Churches NGOs N % N % Pacific 75 20.1 1 6.7 Caribbean 97 26.0 0.0 Andean 179 48.0 14 93.3 Amazon 22 5.9 0.0 Total 373 100.0 15 100.0 Market Population: Description of Churches and NGOs Churches and NGOs by Market Population Ministries were asked to respond to the statement "The location of your ministry headquarters is considered to be: Rural, Suburban, Urban." Ministries local population sizes are summarized in Figure 1.5. The majority of churches and NGOs surveyed were located in urban areas. 18

Figure 1.5 Market Population Size NGOs (N = 15) Local Churches (N = 373) Urban 93.3% 94.9% Suburban 2.1% 6.7% Rural 0.0% 2.9% Location of NGO Work (Local vs. National) Figure 1.6 illustrates the distribution of NGO work from the local to the national level. NGOs were asked Does the work of your ministry in Colombia primarily take place: locally (in one city or in a small number of villages, towns, or neighborhoods), at the department level (in many villages, towns, or cities throughout one department), regionally (in multiple departments in one region), or nationally (in multiple regions of Colombia)? Of the 15 NGOs surveyed, 93.3% reported that they operate nationally and one (6.7%) reported that they operate at the department level. None reported local or regional operation. Figure 1.6 Location of NGO Work (Local vs. National) Department Level 6.7% Nationally 93.3% N = 15 Denominational Affiliation: Description of Churches and NGOs NGOs by Denominational Affiliation Of the 15 NGOs surveyed, 14 (93.3%) identified as Independent, Inter-Denominational, Non- Denominational and one (6.7%) identified as Assemblies of God and other Pentecostal (Apostolic, Foursquare). 19

Churches by Denominational Affiliation Ministries were asked Is your ministry affiliated with a denomination? and If yes, which one? The responses of participating churches are represented in Figure 1.7. Participating churches represented a wide range of denominations. Figure 1.7 Sample Churches by Denominational Affiliation Independent, Inter-Denominational, Non-Denominational Assemblies of God and other Pentecostal (Apostolic, Foursquare) 30.2% 50.4% Baptist, CMA 10.0% Other Evangelical 7.5% Reformed (Presbyterian, Lutheran, Church of Christ) Neo-Pentecostal 1.1% 0.8% N = 371 Churches by Recoded Denominational Affiliation Church denominational affiliation was recoded from six categories to four categories to examine group differences. Reformed (Presbyterian, Lutheran, Church of Christ) and Neo- Pentecostal denominations did not have a sufficient degree of frequency to be used for analysis, therefore, they were not included in further analysis. The distributions of the recoded groups are summarized in Figure 1.8. Figure 1.8 Recoded Church Denominational Affiliation Baptist, CMA 10.2% Independent, Inter- Denominational, Non- Denominational Assemblies of God and other Pentecostal (Apostolic, Foursquare) 30.8% 51.4% Other Evangelical 7.7% N = 364 20

Churches by Recoded Denominational Affiliation and by Recoded Geographic Region Figure 1.9 illustrates the relative number of churches representing each of the four recoded denominational groups within the four recoded geographic regions. In the Andean region, the Independent, Inter-Denominational, Non-Denominational group accounted for 62.3% of church affiliations. In the Amazon region, the Assemblies of God and other Pentecostal (Apostolic, Foursquare) denominational group accounted for over half (54.5%) of church affiliations. In the Pacific region, churches had a more even distribution between the four denominational affiliations. Figure 1.9 Recoded Denominational Affiliation by Recoded Geographic Region Assemblies of God and other Pentecostal (Apostolic, Foursquare) Baptist, CMA Independent, Inter-Denominational, Non-Denominational Other Evangelical Pacific (N = 73) 28.8% 20.5% 32.9% 17.8% Caribbean (N = 94) Andean (N = 175) Amazon (N = 22) 33.0% 27.4% 6.3% 54.5% 8.5% 13.6% 50.0% 62.3% 31.8% 8.5% 4.0% 0.0% Attendance: Description of Churches Churches by Attendance Local churches were asked "Approximately how many people attend a service at your church during a normal week?" The churches surveyed represent a wide range of church sizes, including relatively small, medium, and large churches. The church responses included six attendance ranges from < 25 members to > 500 members attending a service during a normal week. The distribution of these groups is represented in Figure 1.10. 21

Figure 1.10 Local Church Service Attendance 101 200 23.7% 201 500 15.1% > 500 8.9% < 25 1.9% 51 100 35.8% 26 50 14.8% N =372 Churches by Recoded Attendance Church attendance was recoded from six categories to three categories to examine group differences. The categories for < 25 and 26 50 members attending a service during a normal week were combined to form the category 1 50 members. The categories for 51 100 and 101 200 members attending a service during a normal week were combined to form the category 51 200 members. The categories for 201 500 and > 500 members attending a service during a normal week were combined to form the category > 200 members. The distributions of the recoded groups are summarized in Figure 1.11. Figure 1.11 Recoded Local Church Service Attendance 51 200 59.4% > 200 23.9% 1 50 16.7% N = 372 22

Recoded Church Attendance by Recoded Geographic Region There was a statistically significant relationship between recoded church attendance and recoded geographic region. 16 As shown in Figure 1.12, the proportion of churches with > 200 members attending a service during a normal week in the Amazon region was significantly higher than was the proportion of churches in other regions. Likewise, there were no churches with 1 50 members attending a service during a normal week in the Amazon region, while other regions had 15.2% (Andean) to 22.7% (Pacific) of churches represented in this category. Figure 1.12 Recoded Geographic Region by Recoded Church Attendance 1 50 51 200 > 200 Pacific (N = 75) 22.7% 60.0% 17.3% Caribbean (N = 97) 18.6% 57.7% 23.7% Andean (N = 178) 15.2% 61.8% 23.0% Amazon (N = 22) 0.0% 45.5% 54.5% Church Interaction With Children and Youth Ministries were asked Over the past 2 years, has the number of children and youth that your ministry interacts with each year increased by more than 10.0%, increased by 10.0% or less, stayed about the same, decreased by 10.0% or less, or decreased by more than 10.0%? This question was asked separately for children (ages 5 11) and youth (ages 12 19). As shown in Figure 1.13, most churches reported that interactions with children and youth over the past 2 years have increased or stayed about the same. 16 χ 2 (6) = 16.72, p =.010 (highly significant), Cramer's V =.150 (moderate relationship). 23

Figure 1.13 Church Interaction With Children and Youth Over Past 2 Years 60% 40% 20% 0% 37.6% 33.0% Increased by More Than 10.0% 19.1% 18.6% Increased by 10.0% or Less 38.6% 35.8% Stayed About the Same 5.9% 6.8% Decreased by 10.0% or Less Ages 5 11 (N = 372) Ages 12 19 (N = 370) 1.6% 3.0% Decreased by More Than 10.0% Children and Youth in Leadership Roles Ministries were asked Approximately what percentage of the youth (ages 12 19) in your youth ministry have a leadership role in ministry? As shown in Figure 1.14, over half of churches reported that 1.0 10.0% of ages 12 19 hold leadership roles in their churches. Figure 1.14 Percent of Ages 12 19 in Leadership Roles 1.0 10.0% 51.6% 11.0 20.0% 21.0 30.0% > 30.0% 9.9% 16.9% 12.1% None 9.4% N = 372 Church and NGO Age Focus by Ministry Type Ministry Focus Ministries were asked Which of these age groups are an important focus of your ministry: Ages 5 11, 12 15, and/or 16 19? Their responses are represented in Figure 2.1. For local churches, focus appears to decrease as age increases, while for NGOs, focus appears to remain relatively constant. Proportions of local churches and NGOs were similarly focused on ages 5 11. NGOs were significantly more focused on ages 12 15 than were churches. 17 17 χ 2 (1) = 4.18, p =.041 (significant), Cramer's V =.104 (moderate relationship). 24

Figure 2.1 Age Focus of Child and Youth Ministry Categories are not mutually exclusive. Ministries may identify one or more age groups. 120% 100% 91.4% 86.7% 100.0% 86.7% 80% 60% 78.0% 61.9% Local Churches (N = 373) 40% NGOs (N = 15) 20% 0% Ages 5 11 Ages 12 15* Ages 16 19 Note: Categories with statistically significant differences have an asterisk (*). Church Age Focus by Recoded Geographic Region Figure 2.2 summarizes church age focus by recoded geographic region. There were significant relationships between recoded geographic region and age focus for churches. 18 Specifically, significantly greater proportions of churches in the Caribbean and Amazon regions reported focus on ages 5 11 and ages 12 15 than did churches in the Pacific region. A significantly greater proportion of churches in the Caribbean region reported focus on ages 12 15 than did churches in the Andean or Pacific regions. Significantly greater proportions of churches in the Caribbean and Andean regions reported focus on ages 16 19 than did churches in the Pacific and Amazon regions, and the Caribbean proportion was also significantly greater than was the Andean proportion for ages 16 19. 18 Ages 5 11: χ 2 (3) = 11.25, p =.010 (highly significant), Cramer's V =.174 (strong relationship). Ages 12 15: χ 2 (3) = 44.67, p <.001 (highly significant), Cramer's V =.346 (very strong relationship). Ages 16 19: χ 2 (3) = 32.85, p <.001 (highly significant), Cramer's V =.297 (very strong relationship). 25

Figure 2.2 Church Age Focus by Recoded Geographic Region Categories are not mutually exclusive. Ministries may identify one or more age groups. Ages 5 11* Ages 12 15* Ages 16 19* Pacific (N = 75) 53.3% 44.0% 84.0% Caribbean (N = 97) 80.4% 96.9% 93.8% Andean (N = 179) 63.1% 77.7% 90.5% Amazon (N = 22) 31.8% 100.0% 95.5% Note: Categories with statistically significant differences have an asterisk (*). Age Focus by Market Population Church Age Focus Figure 2.3 summarizes church age focus by market population. There were no significant relationships between market population and age focus; however, churches in urban and rural areas reported greater focus on ages 16 19 than did churches in suburban areas (p =.05). Note the small sample sizes of rural and suburban churches. These limit the conclusions that can be drawn about market population differences. Figure 2.3 Local Church Market Population by Age Focus Categories are not mutually exclusive. Ministries may identify one or more age categories. Ages 5 11 Ages 12 15 Ages 16 19 Rural (N = 11) 81.8% 72.7% 100.0% Suburban (N = 8) 25.0% 62.5% 87.5% Urban (N = 354) 62.4% 78.2% 91.2% Note: No statistically significant differences between market populations for any age group. 26

NGO Age Focus Figure 2.4 summarizes NGO age focus by market population. Of the 15 NGOs surveyed, none reported a rural market population, so this population is excluded from the figure. Note that only one NGO reported a suburban market population, and that particular NGO reported focus on all three age groups. NGOs in urban areas reported the greatest focus on ages 12 15, but a high proportion also reported focus on ages 5 11 and ages 16 19. Figure 2.4 NGO Market Population by Age Focus Ages 5 11 Ages 12 15 Ages 16 19 Suburban (N = 1) Urban (N = 14) 85.7% 85.7% 100.0% 100.0% 100.0% 100.0% Note: No statistically significant differences between market populations for any age group. Church Age Focus by Recoded Denominational Affiliation Figure 2.5 illustrates the relationships between recoded denominational affiliations and age focus. Across all denominations, focus appears to decrease as age increases. This is similar to the trend seen among churches overall (see Figure 2.1). There were no significant relationships between recoded denominational affiliation and age focus; however, Assemblies of God and other Pentecostal (Apostolic, Foursquare) churches reported significantly greater focus on ages 12 15 than did Baptist/CMA churches (p =.05). Figure 2.5 Recoded Denominational Affiliation by Age Focus Categories Categories are not mutually exclusive. Ministries may identify one or more age groups. Assemblies of God and other Pentecostal (Apostolic, Foursquare) (N=112) Ages 5 11 Ages 12 15 Ages 16 19 62.5% 94.6% 83.9% Baptist, CMA (N=37) 67.6% 59.5% 97.3% Independent, Inter- Denominational, Non- Denominational (N=187) 61.5% 88.8% 77.0% Other Evangelical (N=28) 85.7% 75.0% 67.9% Note: No statistically significant differences between denominations for any age group. 27

Church Age Focus by Recoded Attendance The relationships between recoded church attendance and age focus are represented in Figure 2.6. There were significant relationships between recoded church attendance and ages 12 15 and 16 19. 19 Significantly greater proportions of churches with > 200 attendees reported focus on ages 12 15 and 16 19 than did all other churches. Further, significantly greater proportions of churches with 51 200 attendees reported focus on ages 16 19 than did churches with 1 50 attendees. Finally, a higher proportion of churches with > 200 attendees reported focus on ages 5 11 than did churches with 51 200 attendees (p =.05). Figure 2.6 Age Focus by Recoded Church Attendance Ages 5 11 Ages 12 15* Ages 16 19* 1 50 (N = 62) 43.5% 66.1% 90.3% 51 200 (N = 221) 60.6% 75.1% 89.6% > 200 (N = 89) 77.5% 96.6% 93.3% Note: Categories with statistically significant differences have an asterisk (*) Programs by Ministry Type Programs (Churches and NGOs) National Statistics for Church Programs The rates of local church programs for ages 5 11 appear in descending order in Figure 3.1A. Churches were most likely to conduct Sunday school (98.7%) and least likely to conduct orphan care (2.1%) for ages 5 11. 19 Ages 12 15: χ 2 (2) = 18.21, p <.001 (highly significant), Cramer's V =.221 (strong relationship). Ages 16 19: χ 2 (2) = 18.21, p <.001 (highly significant), Cramer's V =.221 (strong relationship). 28

Figure 3.1A What Types of Programs Does Your Local Church Conduct for Ages 5 11? Categories are not mutually exclusive. Ministries may identify one or more programs. Sunday School 98.7% Person-to-Person Evangelism 73.2% Bible Clubs or Studies 71.0% Camps or Retreats 70.5% Leadership Training/Development 67.8% Large Evangelistic Gatherings/Festivals 67.6% Scripture Distribution 59.5% Music-Based Events 46.9% Ministry to At-Risk Children and Youth 36.7% Sports Ministry 20.6% School-Based Programs 17.2% Addiction or Trauma Recovery 9.9% Literacy 7.5% N = 373 Orphan Care 2.1% The rates of local church programs for ages 12 19 appear in descending order in Figure 3.1B. Churches were most likely to conduct Sunday school (92.5%) and least likely to conduct orphan care (3.5%) for ages 12 19. 29

Figure 3.1B What Types of Programs Does Your Local Church Conduct for Ages 12 19? Categories are not mutually exclusive. Ministries may identify one or more programs. Sunday School Person-to-Person Evangelism Leadership Training/Development Bible Clubs or Studies Camps or Retreats Large Evangelistic Gatherings/Festivals Scripture Distribution Music-Based Events 92.5% 90.6% 87.9% 82.0% 82.0% 76.7% 70.2% 67.3% Sports Ministry 37.0% Ministry to At-Risk Children and Youth 35.9% School-Based Programs 24.9% Addiction or Trauma Recovery 20.6% Literacy 10.7% Ministry to Youth in Jail or Prison Orphan Care 7.2% 3.5% N = 373 National Statistics for NGO Programs The rates of NGO programs for ages 5 11 appear in descending order in Figure 3.2A. NGOs were most likely to conduct ministry to at-risk children (86.7%) and least likely to conduct orphan care (20.0%) for ages 5 11. 30

Figure 3.2A What Types of Programs Does Your NGO Conduct for Ages 5 11? Categories are not mutually exclusive. Ministries may identify one or more programs. Ministry to At-Risk Children and Youth 86.7% Person-to-Person Evangelism 80.0% Bible Clubs or Studies 73.3% Scripture Distribution 73.3% Large Evangelistic Gatherings/Festivals 73.3% Camps or Retreats 73.3% Leadership Training/Development 73.3% School-Based Programs 66.7% Music-Based Events 53.3% Sports Ministry 53.3% Sunday School 46.7% Addiction or Trauma Recovery 40.0% Literacy Orphan Care 26.7% 20.0% N = 15 The rates of NGO programs for ages 12 19 appear in descending order in Figure 3.2B. NGOs were most likely to conduct leadership training (100%) and least likely to conduct orphan care (26.7%) for ages 12 19. 31

Figure 3.2B What Types of Programs Does Your NGO Conduct for Ages 12 19? Leadership Training/Development 100.0% Large Evangelistic Gatherings/Festivals Camps or Retreats Bible Clubs or Studies Person-to-Person Evangelism Scripture Distribution Ministry to At-Risk Children and Youth School-Based Programs Music-Based Events Sports Ministry Addiction or Trauma Recovery Sunday School Literacy 80.0% 80.0% 73.3% 73.3% 73.3% 73.3% 66.7% 60.0% 53.3% 46.7% 33.3% 26.7% Ministry to Youth in Jail or Prison Orphan Care 26.7% 26.7% N = 15 Church Programs by Recoded Geographic Region Ten of the 14 programs for ages 5 11 were significantly different across recoded geographic regions (see Figure 3.3A). The proportions of churches conducting Bible clubs or studies, personto-person evangelism, and leadership training/development in the Andean region were significantly higher than the proportions of churches in the other regions. 20 There were a significantly higher proportion of churches conducting scripture distribution in the Andean and Amazon regions than in the Pacific region. 21 The proportion of churches conducting large evangelistic gatherings/festivals in the Amazon region was significantly lower than the proportion of churches in the other regions. 22 Music-based events were also significantly less likely in the Amazon and Caribbean regions than in the Andean and Pacific regions. 23 The Caribbean region reported significantly fewer camps or retreats than the other regions. 24 The proportion of churches conducting literacy programs in the Pacific region was significantly higher than in the 20 Bible clubs/studies: χ 2 (3) = 14.92, p =.002 (highly significant), Cramer's V =.200 (strong relationship). Person-to-person evangelism: χ 2 (3) = 9.43, p =.024 (significant), Cramer's V =.159 (strong relationship). Leadership training/development: χ 2 (3) = 17.16, p =.001 (highly significant), Cramer's V =.214 (strong relationship). 21 χ 2 (3) = 11.26, p =.010 (highly significant), Cramer's V =.174 (strong relationship). 22 χ 2 (3) = 11.16, p =.011 (significant), Cramer's V =.173 (strong relationship). 23 χ 2 (3) = 26.36, p <.001 (highly significant), Cramer's V =.266 (very strong relationship). 24 χ 2 (3) = 19.00, p <.001 (highly significant), Cramer's V =.226 (strong relationship). 32

other regions. 25 The Amazon region reported significantly more school-based programs than the other regions. 26 Finally, ministry to at-risk children in the Andean region was significantly lower than in the Pacific and Caribbean regions. 27 Figure 3.3A Statistically Significant Differences in Church Programs by Recoded Geographic Region for Ages 5 11 Categories are not mutually exclusive. Ministries may identify one or more programs. Amazon (N = 22) Andean (N = 179) Caribbean (N = 97) Pacific (N = 75) 31.8% Ministry to At-Risk Children and 25.7% Youth 48.5% 49.3% School-Based Programs Literacy Leadership Training/ Development Camps or Retreats Music-Based Events Large Evangelistic Gatherings/ Festivals Scripture Distribution Person-to-Person Evangelism Bible Clubs or Studies 0.0% 4.5% 7.2% 13.4% 14.4% 18.7% 17.3% 13.6% 33.0% 36.4% Note: Programs shown have statistically significant differences. 48.0% 54.5% 54.5% 52.6% 48.0% 58.8% 58.7% 53.6% 58.1% 70.9% 70.1% 65.3% 62.9% 62.7% 77.3% 78.2% 72.0% 78.2% 72.7% 66.5% 63.6% 80.4% 66.0% 68.0% 59.1% 80.4% 25 χ 2 (3) = 14.61, p =.002 (highly significant), Cramer's V =.198 (strong relationship). 26 χ 2 (3) = 24.03, p <.001 (highly significant), Cramer's V =.254 (very strong relationship). 27 χ 2 (3) = 20.47, p <.001 (highly significant), Cramer's V =.234 (strong relationship). 33

Six of the 15 programs for youth ages 12 19 were significantly different across recoded geographic regions (see Figure 3.3B). Bible clubs or studies were significantly more likely in the Andean region than they were in the Pacific region. 28 A significantly lower proportion of churches conducted large evangelistic gatherings/festivals in the Amazon region than did churches in other regions. 29 A significantly higher proportion of churches conducted literacy programs in the Pacific region than did churches in other regions. 30 A significantly higher proportion of churches conducted school-based programs in the Amazon region than did churches in other regions. 31 Ministry to at-risk youth and recovery from addition or trauma programs were also significantly less likely in the Andean region than they were in the Caribbean and Pacific regions. 32 Figure 3.3B Statistically Significant Differences in Church Programs by Recoded Geographic Region for Ages 12 19 Categories are not mutually exclusive. Ministries may identify one or more programs. Amazon (N = 22) Andean (N = 179) Caribbean (N = 97) Pacific (N = 75) Addiction or Trauma Recovery 27.3% 14.5% 24.7% 28.0% Ministry to At-Risk Children and Youth 31.8% 26.3% 41.2% 53.3% School-Based Programs 17.3% 26.8% 25.3% 77.3% Literacy.0% 8.4% 9.3% 21.3% Large Evangelistic Gatherings/Festivals Bible Clubs or Studies Note: Programs shown have statistically significant differences. 54.5% 74.9% 82.5% 80.0% 77.3% 87.7% 80.4% 72.0% 28 χ 2 (3) = 9.55, p =.023 (significant), Cramer's V =.160 (strong relationship). 29 χ 2 (3) = 8.64, p =.034 (significant), Cramer's V =.152 (strong relationship). 30 χ 2 (3) = 12.70, p =.005 (highly significant), Cramer's V =.185 (strong relationship). 31 χ 2 (3) = 37.93, p <.001 (highly significant), Cramer's V =.319 (very strong relationship). 32 Ministry to at-risk youth: χ 2 (3) = 18.49, p <.001 (highly significant), Cramer's V =.223 (strong relationship). Recovery from trauma or addiction: χ 2 (3) = 8.15, p =.043 (significant), Cramer's V =.148 (moderate relationship). 34

Church Programs by Market Population Three of the 14 programs for ages 5 11 were significantly different across market populations (see Figure 3.4A). 33 A significantly higher proportion of churches conducted leadership training/development programs in urban and suburban areas than did churches in rural areas. Significantly higher proportions of churches in suburban areas conducted school-based and recovery from addiction or trauma programs than did churches in urban areas. Note the small sample sizes of rural and suburban churches. These limit the conclusions that can be drawn about market population differences. Figure 3.4A Statistically Significant Differences in Church Programs for Ages 5 11 by Market Population Rural (N = 11) Suburban (N = 8) Urban (N = 354) Leadership Training/ Development 27.3% 68.4% 100.0% School-Based Programs 18.2% 16.4% 50.0% Addiction or Trauma Recovery 9.0% 18.2% 37.5% Note: Programs shown have statistically significant differences. Two of the 15 programs for youth ages 12 19 were significantly different across market populations (see Figure 3.4B). 34 A significantly higher proportion of churches conducted programs for recovery from addiction or trauma in rural areas than did churches in urban areas. A significantly higher proportion of churches conducted programs for orphan care in suburban areas than did churches in urban areas. Note the small sample sizes of rural and suburban churches. These limit the conclusions that can be drawn about market population differences. Figure 3.4B Statistically Significant Differences in Church Programs for Ages 12 19 by Market Population Rural (N = 11) Suburban (N = 8) Urban (N = 354) Addiction or Trauma Recovery 25.0% 19.5% 54.5% Orphan Care 0.0% 3.1% 25.0% Note: Programs shown have statistically significant differences. 33 Leadership training/development: χ 2 (2) = 12.13, p =.002 (highly significant), Cramer's V =.180 (strong relationship). School-based programs: χ 2 (2) = 6.23, p =.044 (significant), Cramer's V =.129 (moderate relationship). Recovery from addiction or trauma: χ 2 (2) = 7.96, p =.019 (significant), Cramer's V =.146 (moderate relationship). 34 Recovery from addiction or trauma: χ 2 (2) = 8.10, p =.017 (significant), Cramer's V =.147 (moderate relationship). Orphan care: χ 2 (2) = 11.56, p =.003 (highly significant), Cramer's V =.176 (strong relationship). 35

Church Programs by Recoded Attendance Four of the 15 programs for ages 5 11 showed significant variation across recoded church attendance categories (see Figure 3.5A). 35 A significantly higher proportion of churches with more than 200 attendees conducted scripture distribution than churches with 1-50 or 51-200 attendees. Significantly higher proportions of churches with51-200 and more than 200 attendees conducted music-based events than churches with 1-50 attendees. Churches with more than 200 attendees were significantly more likely to conduct camps and retreats than churches with 51-200 attendees, and these both were significantly more likely to conduct camps and retreats than churches with 1-50 attendees. Finally, a significantly higher proportion of churches with more than 200 attendees conducted school-based programs than churches with 51-200 attendees. Sunday school results indicated significance; however, the test results were not reliable because of the imbalance in sample sizes caused by the very high program rates across church attendance categories. Figure 3.5A Statistically Significant Differences for Ages 5 11 in Church Programs by Recoded Attendance 1 50 (N = 62) 51 200 (N = 221) > 200 (N = 89) Scripture Distribution 48.4% 57.9% 70.8% Music Based Events 32.3% 48.4% 52.8% 50.0% Camps or Retreats 70.1% 85.4% School-Based Programs 16.1% 13.6% 25.8% Note: Programs shown have statistically significant differences. 35 Scripture distribution: χ 2 (2) = 8.10, p =.017 (significant), Cramer's V =.148 (moderate relationship). Music-based events: χ 2 (2) = 6.79, p =.034 (significant), Cramer's V =.135 (moderate relationship). Camps or retreats: χ 2 (2) = 22.00, p <.001 (highly significant), Cramer's V =.243 (strong relationship). School-based programs: χ 2 (2) = 6.82, p =.033 (significant), Cramer's V =.135 (moderate relationship). 36

Nine of the 15 programs for ages 12 19 showed significant variation across recoded church attendance categories (see Figure 3.5B). 36 Generally, programs were conducted at higher rates in churches with higher attendance (more than 200 attendees) than at churches with lower attendance (1-50 attendees). Significantly higher proportions of churches with 51-200 and more than 200 attendees conducted Sunday school and person-to-person evangelism than churches with 1-50 attendees. Similarly, significantly higher proportions of churches with more than 200 attendees conducted large evangelistic gatherings/festivals and school-based programs than churches with 1-50 or 51-200 attendees. Scripture distribution was significantly more likely in churches with more than 200 attendees than churches with 51-200 attendees. Finally, churches with more than 200 attendees were significantly more likely to conduct Bible clubs or studies, music-based events, camps or retreats, and leadership training/development than churches with 51-200 attendees, and these both were significantly more likely to conduct these programs than churches with 1-50 attendees. 36 Sunday school: χ 2 (2) = 11.33, p =.003 (highly significant), Cramer's V =.175 (strong relationship). Bible clubs or studies: χ 2 (2) = 13.04, p =.001 (highly significant), Cramer's V =.187 (strong relationship). Person-to-person evangelism: χ 2 (2) = 11.75, p =.003 (highly significant), Cramer's V =.178 (strong relationship). Scripture distribution: χ 2 (2) = 6.48, p =.039 (significant), Cramer's V =.132 (moderate relationship). Large evangelistic gatherings/festivals: χ 2 (2) = 13.27, p =.001 (highly significant), Cramer's V =.189 (strong relationship). Music-based events: χ 2 (2) = 22.54, p <.001 (highly significant), Cramer's V =.246 (strong relationship). Camps or retreats: χ 2 (2) = 30.40, p <.001 (highly significant), Cramer's V =.286 (very strong relationship). Leadership training/development: χ 2 (2) = 14.91, p =.001 (highly significant), Cramer's V =.200 (strong relationship). School-based programs: χ 2 (2) = 23.66, p <.001 (highly significant), Cramer's V =.252 (very strong relationship). 37

Figure 3.5B Statistically Significant Differences for Ages 12 19 in Church Programs by Recoded Attendance Categories are not mutually exclusive. Ministries may identify one or more programs. 1 50 (N = 62) 51 200 (N = 221) > 200 (N = 89) Sunday School Bible Clubs or Studies 82.3% 94.1% 95.5% 69.4% 81.9% 92.1% Person-to-Person Evangelism 79.0% 93.2% 92.1% Scripture Distribution 67.7% 66.5% 80.9% Large Evangelistic Gatherings/Festivals 66.1% 74.2% 89.9% Music-Based Events 45.2% 67.4% 82.0% Camps or Retreats 59.7% 83.3% 94.4% Leadership Training/ Development 75.8% 87.8% 96.6% School-Based Programs 14.5% 19.9% 43.8% Note: Programs shown have statistically significant differences. 38

Program Categories Recoded Program Categories Each program for children and youth was assigned to one of five categories: evangelism, discipleship, benevolence, evangelism and discipleship, or evangelism and benevolence. (see Figure 4.1). Figure 4.1 Program Categories Programs for Children and Youth Person-to-Person Evangelism Scripture Distribution Large Evangelistic Gatherings/Festivals Leadership Training/Development Literacy Music-Based Events Orphan Care Sunday School Bible Clubs or Studies Camps or Retreats Sports Ministry School-Based Programs Ministry to At-Risk Children and Youth Ministry to Youth in Jail or Prison Addiction or Trauma Recovery Categories Evangelism Discipleship Benevolence Evangelism and/or Discipleship Evangelism and/or Benevolence Program Categories by Ministry Type Church Program Categories The percentages of local churches conducting specific program categories appear in Figure 4.2. Almost all churches conducted evangelism and/or discipleship programs for ages 5 11 (95.4%) and ages 12 19 (94.8%); however, they were less likely to conduct benevolence programs for ages 5 11 (52.0%) and ages 12 19 (74.5%). 39

Figure 4.2 Program Categories for Local Churches Evangelism Programs for Ages 5 11 90.3% Discipleship Programs for Ages 5 11 67.8% Benevolence Programs for Ages 5 11 49.3% Evangelism/Discipleship Programs for Ages 5 11 100.0% Evangelism/Benevolence Programs for Ages 5 11 45.3% Evangelism Programs for Ages 12 19 96.0% Discipleship Programs for Ages 12 19 87.9% Benevolence Programs for Ages 12 19 69.4% Evangelism/Discipleship Programs for Ages 12 19 98.4% Evangelism/Benevolence Programs for Ages 12 19 53.9% N = 373 NGO Program Categories The percentages of NGOs conducting specific program categories appear in Figure 4.3. Like local churches, almost all NGOs conducted evangelism and/or discipleship programs for ages 5 11 (93.3%) and ages 12 19 (100.0%), and they were less likely to conduct benevolence programs for ages 5 11 (66.7%) and ages 12 19 (73.3%). 40

Figure 4.3 Program Categories for NGOs Evangelism Programs for Ages 5 11 Discipleship Programs for Ages 5 11 Benevolence Programs for Ages 5 11 Evangelism/Discipleship Programs for Ages 5 11 Evangelism/Benevolence Programs for Ages 5 11 Evangelism Programs for Ages 12 19 Discipleship Programs for Ages 12 19 Benevolence Programs for Ages 12 19 Evangelism/Discipleship Programs for Ages 12 19 Evangelism/Benevolence Programs for Ages 12 19 93.3% 73.3% 66.7% 93.3% 93.3% 93.3% 100.0% 73.3% 100.0% 93.3% N = 15 Church Program Categories by Recoded Denominational Affiliation Figure 4.4 illustrates the relationships between church program categories and recoded denominational affiliations. There was a statistically significant relationship between evangelism/benevolence programs and the recoded denominational affiliation for ages 12 19. The proportion of Other Evangelical churches conducting these programs was significantly higher than was the proportion of Independent, Inter-Denominational, Non-Denominational and Baptist/CMA churches. Further, the proportion of Assemblies of God and Other Pentecostal (Apostolic, Foursquare) churches was significantly higher than was the proportion of Independent, Inter-Denominational, Non-Denominational churches. 37 There were not statistically significant relationships between church program categories and recoded denominational affiliation for ages 5 11. However, the proportion of Assemblies of God and other Pentecostal (Apostolic, Foursquare) churches conducting benevolence programs for ages 5 11 was higher than was the proportion of Independent, Inter-Denominational, Non-Denominational churches (p =.05). 37 χ 2 (3) = 16.99, p =.001 (highly significant), Cramer's V =.217 (strong relationship). 41

Figure 4.4 Program Categories by Recoded Denominational Affiliation for Churches Assemblies of God and Other Pentecostal (Apostolic, Foursquare) (N = 112) Baptist, CMA (N = 37) Independent, Inter-Denominational, Non-Denominational (N = 187) Other Evangelical (N = 28) Evangelism Programs for Ages 5 11 Discipleship Programs for Ages 5 11 Benevolence Programs for Ages 5 11 Evangelism/Discipleship Programs for Ages 5 11 Evangelism/Benevolence Programs for Ages 5 11 Evangelism Programs for Ages 12 19 Discipleship Programs for Ages 12 19 Benevolence Programs for Ages 12 19 Evangelism/Discipleship Programs for Ages 12 19 Evangelism/Benevolence Programs for Ages 12 19* 52.7% 51.4% 39.0% 50.0% 71.4% 73.0% 65.2% 57.1% 58.9% 43.2% 46.0% 42.9% 62.5% 54.1% 44.9% Note: Categories with statistically significant differences have an asterisk (*). 76.8% 64.9% 66.3% 67.9% 92.0% 86.5% 89.8% 92.9% 100.0% 100.0% 100.0% 100.0% 98.2% 94.6% 95.7% 92.9% 92.0% 86.5% 86.1% 82.1% 78.6% 99.1% 97.3% 97.9% 100.0% 42

Church Program Categories by Recoded Geographic Region Figure 4.5 represents the relationships between church program categories and recoded geographic regions. There were significant relationships between the recoded geographic region and: discipleship programs for ages 5 11, benevolence programs for ages 5 11, evangelism and/or benevolence programs for ages 5 11, and evangelism and/or benevolence programs for ages 12 19. Significantly more churches conducted discipleship programs for ages 5 11in the Andean region than did in other regions. 38 Significantly more churches conducted benevolence programs for ages 5 11in the Andean and Pacific regions than did in Caribbean and Amazon regions. 39 Proportions of local churches conducting evangelism and/or benevolence programs for ages 5 11 were significantly different among Amazon, Caribbean, and Andean regions; the Amazon region had the highest proportion, and the Andean region had the lowest proportion. The Pacific region also had significantly higher proportions of churches conducting evangelism and/or benevolence programs for ages 5 11 than did the Andean region. 40 Proportions of local churches conducting evangelism and/or benevolence programs for ages 12 19 were significantly different among all four regions; the Amazon region had the highest proportion, and the Andean region had the lowest proportion. 41 38 χ 2 (3) = 17.16, p =.001 (highly significant), Cramer's V =.214 (strong relationship). 39 χ 2 (3) = 25.51, p <.001 (highly significant), Cramer's V =.261 (very strong relationship). 40 χ 2 (3) = 29.64, p <.001 (highly significant), Cramer's V =.282 (very strong relationship). 41 χ 2 (3) = 47.64, p <.001 (highly significant), Cramer's V =.358 (very strong relationship). 43

Figure 4.5 Program Categories by Recoded Geographic Region for Churches Pacific (N = 75) Caribbean (N = 97) Andean (N = 179) Amazon (N = 22) Evangelism Programs for Ages 5 11 Discipleship Programs for Ages 5 11* 58.7% 58.8% 54.5% 89.3% 85.6% 93.3% 90.9% 78.2% Benevolence Programs for Ages 5 11* 13.6% 36.1% 53.3% 59.2% Evangelism/Discipleship Programs for Ages 5 11 100.0% 100.0% 100.0% 100.0% Evangelism/Benevolence Programs for Ages 5 11* 31.8% 58.7% 52.6% 77.3% Evangelism Programs for Ages 12 19 Discipleship Programs for Ages 12 19 Benevolence Programs for Ages 12 19 Evangelism/Discipleship Programs for Ages 12 19 94.7% 96.9% 95.5% 100.0% 90.7% 85.6% 87.7% 90.9% 69.3% 61.9% 74.3% 63.6% 98.7% 99.0% 97.8% 100.0% Evangelism/Benevolence Programs for Ages 12 19* 38.0% 57.7% 74.7% 95.5% 44

Categories of Church Programs by Market Population The relationships between church program categories and market population are represented in Figure 4.6. A statistically significant relationship was found between discipleship programs for ages 5 11 and market population. The proportion of churches in urban and suburban areas conducting discipleship programs for ages 5 11 was significantly higher than were the proportion of churches in rural areas. 42 Note the small sample sizes of rural and suburban churches. These limit the conclusions that can be drawn about market population differences. Figure 4.6 Program Categories by Market Population for Churches Rural (N = 11) Suburban (N = 8) Urban (N = 354) Evangelism Programs for Ages 5 11 90.9% 87.5% 90.4% Discipleship Programs for Ages 5 11* 27.3% 68.4% 100.0% Benevolence Programs for Ages 5 11 36.4% 37.5% 50.0% Evangelism/Discipleship Programs for Ages 5 11 100.0% 100.0% 100.0% Evangelism/Benevolence Programs for Ages 5 11 45.5% 44.9% 62.5% Evangelism Programs for Ages 12 19 Discipleship Programs for Ages 12 19 Benevolence Programs for Ages 12 19 Evangelism/Discipleship Programs for Ages 12 19 Evangelism/Benevolence Programs for Ages 12 19 100.0% 100.0% 95.8% 72.7% 100.0% 88.1% 81.8% 75.0% 68.9% 100.0% 100.0% 98.3% 63.6% 62.5% 53.4% 42 χ 2 (2) = 12.13, p =.002 (highly significant), Cramer's V =.180 (strong relationship). 45

Church Program Categories by Recoded Attendance Figure 4.7 displays the relationships between church program categories and recoded attendance. There were statistically significant relationships between recoded church attendance and: evangelism programs for ages 5 11, evangelism programs for ages 12 19, discipleship programs for ages 12 19, benevolence programs for ages 12 19, evangelism/discipleship programs for ages 12 19, and evangelism/benevolence programs for ages 12 19. As shown, all program category proportions were higher for churches with > 200 attendees than were the proportions for churches with fewer attendees. Specifically, the proportions of churches with 51 200 and > 200 attendees conducting evangelism programs for ages 5 11 and 12 19 were significantly higher than were the proportions of churches with 1 50 attendees; the proportions of churches with 51 200 and > 200 attendees conducting evangelism/discipleship programs for ages 12 19 were significantly higher than were the proportions of churches with 1 50 attendees. 43 Proportions of local churches conducting discipleship programs for ages 12 19 and benevolence programs for ages 12 19 were significantly different among all three attendance categories, churches with > 200 attendees had the highest proportion, and churches with 1 50 attendees had the lowest proportion. 44 The proportion of local churches with > 200 attendees conducting evangelism/benevolence programs for ages 12 19 was significantly higher than was the proportion of churches with 1 50 or 51 200 attendees. 45 43 Evangelism programs for ages 5 11: χ 2 (2) = 12.07, p =.002 (highly significant), Cramer's V =.180 (strong relationship). Evangelism programs for ages 12 19: χ 2 (2) = 11.03, p =.004 (highly significant), Cramer's V =.173 (strong relationship). Evangelism/discipleship for ages 12 19: χ 2 (2) = 11.48, p =.003 (highly significant), Cramer's V =.176 (strong relationship). 44 Discipleship programs for ages 12 19: χ 2 (2) = 15.26, p =.001 (highly significant), Cramer's V =.203 (strong relationship). Benevolence programs for ages 12 19: χ 2 (2) = 18.35, p <.001 (highly significant), Cramer's V =.223 (strong relationship). 45 χ 2 (2) = 13.18, p =.001 (highly significant), Cramer's V =.189 (strong relationship). 46

Figure 4.7 Program Categories for Children and Youth by Church Size 1 50 (N = 62) 51 200 (N = 221) > 200 (N = 89) Evangelism Programs for Ages 5 11* Discipleship Programs for Ages 5 11 Benevolence Programs for Ages 5 11 Evangelism/Discipleship Programs for Ages 5 11 Evangelism/Benevolence Programs for Ages 5 11 Evangelism Programs for Ages 12 19* Discipleship Programs for Ages 12 19* 79.0% 91.4% 95.5% 58.1% 67.9% 74.2% 37.1% 50.2% 55.1% 100.0% 100.0% 100.0% 37.1% 44.3% 52.8% 88.7% 96.8% 98.9% 75.8% 87.8% 96.6% Benevolence Programs for Ages 12 19* Evangelism/Discipleship Programs for Ages 12 19* 50.0% 69.7% 82.0% 93.5% 99.1% 100.0% Evangelism/Benevolence Programs for Ages 12 19* 38.7% 52.5% 67.4% Note: Programs with statistically significant differences have an asterisk (*). 47

Frequency of Technology Use by Churches Technology Ministries were asked How often does your ministry use each of the following digital technologies to minister to or communicate with youth? The frequency of technology use by churches appears in Figure 5.1. Churches were most likely to report use of video (52.5%) and social media (51.7%) but least likely to use websites (51.8%). Figure 5.1 How Often Are the Following Technologies Used for the Ministry? Never Sometimes Often We Plan to Begin in the Next Year Texting (N = 373) 16.6% 41.6% 39.9% 1.9% Email (N = 373) 12.6% 46.1% 39.9% 1.3% Social Media (N = 373) 20.9% 25.5% 51.7% 1.9% Website (N = 371) 51.8% 24.0% 18.6% 5.7% Video (N = 373) 11.5% 35.4% 52.5% 0.5% Frequency of Technology Use by Ages 12 19 in Churches Ministries were asked How many youth in your ministry area have a cell phone, a smart phone, easy internet access, or equipment to watch videos or DVDs at home? The frequency of technology use by ages 12 19 appears in Figure 5.2. Large proportions of youth ages 12 19 were reported to have cell phones, easy internet access, and video equipment. Youth ages 12 19 were much less likely to have smart phones. Figure 5.2 How Many Youth (Ages 12 19) Have the Following Technologies? None About 25.0% About 50.0% About 75.0% All 4.1% Cellphones (N = 370) 8.4% 38.6% 48.9% Smartphones (N = 373) Internet (N = 373) Video (N = 371) 1.1% 7.2% 0.8% 10.8% 23.6% 14.7% 11.1% 26.5% 23.5% 53.6% 50.4% 53.9% 15.8% 6.4% 0.5% 48

Secondary Impact by Ministry Type Secondary Impact Church Programs That May Impact Children and Youth Ministries were asked "What types of programs or activities does your ministry conduct that may impact children and youth?" In Figure 6.1, the percentages of specific programs conducted by churches that may impact children and youth appear in descending order. Churches were most likely to conduct food programs (41.6%) and least likely to conduct water or sanitation programs (5.6%). Figure 6.1 What Types of Programs Does Your Local Church Conduct That May Impact Children and Youth? Categories are not mutually exclusive. Ministries may identify one or more programs. Food Programs 41.6% Health Programs 31.6% Life Skills Training 25.5% Water or Sanitation Programs 5.6% N = 373 NGO Programs That May Impact Children and Youth In Figure 6.2, the percentages of specific programs conducted by NGOs that may impact children and youth appear in descending order. NGOs were also most likely to conduct food programs (53.3%) and least likely to conduct water or sanitation programs (20.0%). Figure 6.2 What Types of Programs Does Your NGO Conduct That May Impact Children and Youth? Categories are not mutually exclusive. Ministries may identify one or more programs. Food Programs 53.3% Life Skills Training 46.7% Health Programs 33.3% Water or Sanitation Programs 20.0% N = 15 49

Local Church Programs That May Impact Children and Youth by Market Population Size The relationships between church programs that may impact children and youth and market population size are displayed in Figure 6.3. One of the four programs that may impact children and youth was significantly different across market population sizes. The proportion of churches conducting life skills training in urban areas was significantly lower than were the proportion of churches in suburban areas. 46 Note the small sample sizes of rural and suburban churches. These limit the conclusions that can be drawn about market population differences. Figure 6.3 Local Church Programs That May Impact Children and Youth by Market Population Size Categories are not mutually exclusive. Ministries may identify one or more programs. Urban (N = 354) Suburban (N = 8) Rural (N = 11) Life Skills Training* 24.3% 36.4% 62.5% Health Programs 31.4% 27.3% 50.0% Water or Sanitation Programs 5.4% 12.5% 9.1% Food Programs 27.3% 41.5% 62.5% Note: Programs with statistically significant differences have an asterisk (*). Local Church Programs That May Impact Children and Youth by Recoded Geographic Region The relationships between church programs that may impact children and youth and recoded geographic region are displayed in Figure 6.4. One of the four programs which may impact children and youth was significantly different across recoded geographic regions. The proportion of churches conducting life skills training in the Amazon region was significantly higher than was the proportion of churches in the Caribbean and Andean regions. Further, the Pacific region had a significantly higher proportion of churches conducting life skills training than did the Andean region. 47 46 χ 2 (2) = 6.73, p =.035 (significant), Cramer's V =.134 (moderate relationship). 47 χ 2 (3) = 11.97, p =.008 (highly significant), Cramer's V =.179 (strong relationship). 50

Figure 6.4 Local Church Programs That May Impact Children and Youth by Recoded Geographic Region Amazon (N = 22) Andean (N = 179) Caribbean (N = 97) Pacific (N = 75) Life Skills Training* 19.6% 25.8% 32.0% 50.0% Health Programs 27.3% 26.8% 34.0% 41.3% Water or Sanitation Programs 0.0% 4.5% 4.0% 10.3% Food Programs 37.4% 45.5% 44.3% 46.7% Local Church Programs That May Impact Children and Youth by Recoded Denominational Affiliation The relationships between church programs that may impact children and youth and recoded denominational affiliation are displayed in Figure 6.5. All four programs which may impact children and youth were significantly different across recoded denominational affiliations. The proportions of Assemblies of God and other Pentecostal (Apostolic, Foursquare) and Baptist/CMA churches conducting life skills training, health programs, water or sanitation programs, and food programs were significantly higher than were the proportions of for Independent, Inter-Denominational, and Non-Denominational churches. 48 48Life skills training: χ 2 (3) = 11.12, p =.011 (significant), Cramer's V =.175 (strong relationship). Health programs: χ 2 (3) = 8.42, p =.038 (significant), Cramer's V =.152 (strong relationship). Water or sanitation programs: χ 2 (3) = 12.25, p =.007 (highly significant), Cramer's V =.183 (strong relationship). Food programs: χ 2 (3) = 11.47, p =.009 (highly significant), Cramer's V =.178 (strong relationship). 51

Figure 6.5 Local Church Programs That May Impact Youth by Recoded Denominational Affiliation Other Evangelical (N = 28) Independent, Inter-denominational, Non-Denominational (N = 187) Baptist, CMA (N = 37) Assemblies of God and other Pentecostal (Apostolic, Foursquare) (N = 112) Life Skills Training* 17.9% 19.8% 37.8% 33.9% Health Programs* 25.0% 26.2% 43.2% 39.3% Water or Sanitation Programs* 3.6% 2.1% 13.5% 9.8% Food Programs* 35.7% 34.2% 48.2% 59.5% Reach Reported Direct and Indirect Impact for Churches and NGOs Reported Impact or Engagement is a respondent s indicated annual engagement with children and youth aged 5 19. How many children and youth does a ministry impact annually? Of the ministries surveyed, all churches (372) and NGOs (16) responded to the question "Approximately how many children (ages 5 11) and youth (ages 12 19) does your ministry impact each year?" 49 On average, local churches reported reaching approximately 59 children (SD = 160) and 64 youth (SD = 424) each year. NGOs reported reaching an average of 14,728 (SD = 25,169) children and 11,419 (SD = 15,710) youth each year. Reports varied widely from a minimum of 5 to a maximum of 150,000 children and youth reached per ministry. Half of the local churches "numbers reached" were above 50 and half were below, representing the median. For NGOs, the median value was 8,000. The data calculating ministries reach are summarized in Figures 7.1 (annual direct reach for churches) and 7.2 (annual direct reach for NGOs) for children (5 11) and youth (12 19) separately. Overall totals across children and youth are 49 One ministry that self-identified as a church was included as an NGO in the reach calculations based on investigation of the website. 52

presented in the text in the next section. Churches: o Ages 5-11: 22,027 o Ages 12-19: 23,737 o Total 50 : 45,708 NGOs: o Ages 5-11: 235,652 o Ages 12-19: 182,707 o Total 51 : 418,379 Total: o Ages 5-11: 257,679 o Ages 12-19: 206,444 o Total 52 : 464,087 Calculating Annual Direct Gospel Reach for Churches and NGOs Gospel Reach After being asked about the number of children and youth that the ministry engages or interacts with, they were asked Of those children (ages 5 11) and/or youth (ages 12 19) numbered in previous question, to approximately how many does your ministry present the Gospel of salvation through Jesus Christ? Enter the actual number or your best estimate. Count each individual only once. After adjustment for Gospel content, the average number of children and youth reached by local churches remained about the same at 120 (SD = 570). The average number of children and youth reached by NGOs decreased to 13,562 (SD = 19,486). Median values stayed at 50 for local churches and decreased to 4,345 for NGOs. Direct vs. Indirect Reach After being asked about Gospel Reach, they were asked about the number of children they have direct personal interaction with. Specifically they were asked, Of those children (ages 5 11) and/or youth (ages 12 19) numbered in the previous question, how many are presented with the Gospel in a direct personal interaction where the child or youth can respond or ask questions or pray with the presenter or a ministry representative? After adjustment for indirect reach, the average number of children and youth reached by local churches decreased to 105 (SD = 565). The average number of children and youth reached by NGOs decreased to 8,723 (SD = 12,463). Median values were 42 for local churches and 3,100 for NGOs. Length of Ministry Engagement Responses to another survey question were used to further adjust the number of children and youth reached based on the average length of time the ministry works with them. The survey asked, "Approximately how long does your ministry work with each individual child or youth, on average?" with response options of "Less than a year", "1 year but less than 2 years", "2 years but less than 3 years", or "3 years or more". Where the ministry indicated "3 years or more", its numbers reached were reduced to one third of the previously adjusted counts; if the ministry indicated "2 years but less than 3 years", its numbers reached were reduced to one half of the previously adjusted counts. This adjustment seeks to reduce duplicated counting where ministries work with 50 The totals do not equal the sum of the older and younger children because some organizations only answered the total but did not provide any information about the breakdown between older and younger children. 51 Same as note 24. 52 Same as note 24. 53

the same children for multiple years. This question was asked separately for children (5 11 years) and youth (12 19 years). The average number of children reached was 19 (SD = 54, Median = 10) and the average number of youth reached was 12 (SD = 142, Median = 8) for local churches. For NGOs, the average number of children reached was 4,049 (SD = 7,889, Median = 851) and the average number of youth reached was 1,883 (SD = 2,351, Median = 638). NGO Ministry Through Churches Finally, NGOs were also asked what proportion of their ministry to children and youth occurred through local churches. The answer options included none, a small amount, about half, the majority, and all. If the NGO indicated that all of their reach occurred through local churches, then their total was adjusted to 0. If they said the majority occurred through local churches, their total was adjusted down to 25.0% of the previous total. If they said about half then their reach was cut in half. If they said a small amount occurred through local churches then 75.0% of their reach was kept. Finally, if none of their reach occurred through local churches, their total was not adjusted. This was done to account for overlap between the reported reach of the local churches and the NGOs. After adjustments, the average number of children reached was 2,931 (SD = 7,982, Median = 203) and the average number of youth reached was 955 (SD = 1,775, Median = 193). Figures 7.1 and 7.2 show the annual reach for churches and NGOs. Figure 7.1 Calculating Annual Direct Gospel Reach for Churches Child (5 to 11 years) Youth (12 to 19 years) N Annual Engagement Number Gospel Presented to Engaged Gospel Directly Presented to Engaged Length of Ministry Engagement Valid 372 372 372 372 Missing 0 0 0 0 Mean 59 58 48 19 Median 27 27 20 10 Mode 20 20 10 5 Std. Deviation 160 160 156 54 Minimum 0 0 0 0 Maximum 2,500 2,500 2,500 825 Sum 22,027 21,537 17,842 7,160 N Valid 372 372 372 372 Missing 0 0 0 0 Mean 64 62 57 23 Median 21 20 19 8 Mode 10 15 15 7 Std. Deviation 424 424 421 142 Minimum 0 0 0 0 Maximum 8,000 8,000 8,000 2,640 Sum 23,737 23,245 21,137 8,537 54

Figure 7.2 Calculating Annual Direct Gospel Reach for NGOs Child (5 to 11 years) Youth (12 to 19 years) N Annual Engagement Number Gospel Presented to Engaged Gospel Directly Presented to Engaged Length of Ministry Engagement Adjusted by Amount of Ministry through Churches Valid 16 14 12 12 12 Missing 0 2 4 4 4 Mean 14,728 6,979 5,507 4,049 2,931 Median 5,300 2,920 1,860 851 203 Mode 120 1,000 120 83 0 Std. Deviation 25,169 10,905 8,637 7,889 7,982 Minimum 120 120 120 83 0 Maximum 100,000 34,000 28,050 28,050 28,050 Sum 235,652 97,710 66,080 48,588 35,169 N Valid 16 14 12 12 12 Missing 0 2 4 4 4 Mean 11,419 6,584 3,217 1,883 955 Median 3,200 2,225 1,340 638 193 Mode 2400 800 200 66 0 Std. Deviation 15,710 11,399 5,007 2,351 1,775 Minimum 400 400 200 66 0 Maximum 50,000 36,000 18,000 6,000 4,950 Sum 182,707 92,170 38,600 22,601 11,460 Annual Direct Gospel Reach for Churches and NGOs Churches: o Ages 5-11: 7,160 o Ages 12-19: 8.537 o Total: 15,698 NGOs: o Ages 5-11: 35,169 o Ages 12-19: 11,460 o Total: 46,630 Total: o Ages 5-11: 42,330 o Ages 12-19: 19,998 o Total: 62,327 55

Annual Church Reach by Recoded Geographic Region Figures 7.3 through 7.6 show the breakdowns of annual reach by recoded geographic region, recoded denominational affiliation, and recoded attendance for churches. Both the total annual reach and the average annual reach per church are shown. Specifically, the total reach shows how the reach was broken down based on the specific sample for this project. Regional areas and denominations that made up more of the sample showed more reach. The average annual reach, in contrast, presents the average reach of all the churches in that region, denomination, or church size to facilitate making comparisons between groups. Annual Church Reach by Recoded Geographic Region Figure 7.3 Total Annual Church Reach by Recoded Geographic Region Amazon (N=22) Andean (N=179) Caribbean (N=97) 122 311 Ages 5 11 Ages 12 19 2,818 2,641 1,034 1,085 Pacific (N=75) 2,998 4,689 Figure 7.4 Average Annual Church Reach by Recoded Geographic Region Ages 5 11 Ages 12 19 Amazon (N=22) Andean (N=179) Caribbean (N=97) 5.57 14.12 15.83 14.84 10.65 11.19 Pacific (N=75) 39.97 62.52 56

Annual Church Reach by Recoded Denominational Affiliation Figure 7.5 Total Annual Church Reach by Recoded Denominational Affiliation Ages 5 11 Ages 12 19 Assemblies of God and Other Pentecostal (Apostolic, Foursquare) (N=111) 1,640 2,305 Baptist, CMA (N=37) 424 561 Independent, Inter- Denominational, Non-Denominational (N=187) 2,968 3,399 Other Evangelical (N=28) 1,243 2,983 Figure 7.6 Average Annual Church Reach by Recoded Denominational Affiliation Ages 5 11 Ages 12 19 Assemblies of God and Other Pentecostal (Apostolic, Foursquare) (N=112) Baptist, CMA (N=37) Independent, Inter- Denominational, Non-Denominational (N=187) 20.77 14.77 15.17 11.46 15.87 18.18 Other Evangelical (N=28) 44.39 106.53 57

Annual Church Reach by Recoded Attendance Figure 7.7 Total Annual Church Reach by Recoded Attendance Ages 5 11 Ages 12 19 1 50 (N = 62) 563 443 51 200 (N = 221) 2,392 2,827 > 200 (N = 89) 3,771 5,703 Figure 7.8 Average Annual Church Reach by Recoded Attendance Ages 5 11 Ages 12 19 1 50 (N = 62) 51 200 (N = 221) 9.08 7.14 12.79 10.82 > 200 (N = 89) 42.37 64.08 Calculated National Reach for Churches and NGOs Calculated National Reach was arrived at by multiplying median church Reach (18) by the number of Protestant and evangelical churches reported by Operation World (20,043) and adding that total (360,774) to total NGO Annual Direct Gospel Reach (46,630) for a final 407404 53. Churches median (18) x protestant churches in Colombia (20,043) = churches total (360,774) NGOs sum (46,630) + churches total (360,774) = National Total (407404) 53 Reach calculations should be considered as preliminary and are subject to further adjustment based on such factors as incorporating actual mega-church reach and further clarification from participating ministries. 58

LIMITATIONS There is potential overlap in the Reach estimate for NGOs and churches, which would tend to inflate the national Reach estimate. Churches and NGOs may reach the same children and youth, either incidentally, or by intentionally working in direct partnership. This would be especially common in countries with a large Christian population. The church external Reach data were adjusted down based on several factors related to OneHope s internal missional Reach definition (see Methods section). In Colombia, the majority of churches and NGOs surveyed were located in urban areas. As a result, ministry activity in rural areas is under-represented in this study. 59

ACKNOWLEDGEMENTS Colombia Project Team Astrid Cortes served as National Project Coordinator and led the in-country effort for this project. Also serving the in-country effort: Leila Romero, Jair Pencue y Luis Cardenas Andean Region (Bogota), Omar Bohorquez Andean Region (Santander), Gustavo Echeverry Andean Region (Antioquia), Magdalena Betancourt Caribbean Region, Jorge Betancurt & Aura Isabel Granada Pacific Region, and Leila Guzman Amazon & Orinoquia Region. Josh Parker, OneHope International Project Coordinators, led from the home office. Special thanks for cooperation, input, and expertise: Ricardo Luna, OneHope Regional Director, Latin America. Special thanks also to Willington Ortiz, OneHope Director for Colombia, who provided support and d in country contacts for this project, and to Ricardo Luna, OneHope Regional Director, Latin America. Analysis and Content Team Jaquelyn Pennings, Ph.D., Pstat, Director of Non-Profit Research, and an accredited professional statistician for Elite Research, LLC, assisted with survey development and data analysis. Kathleen Coenen, M.S.W, independent contractor, assisted with data analysis. Christine Schaeffer, M.B.A., Research Consultant, assisted with report preparation. Ceri Usmar, Editor, assisted with report editing. Advisory Feedback and Survey Development Panel Tena Stone, PhD, Director of Research and Training, OneHope, managed the advisory feedback process and contributed to survey development. Patricia McClung, Ed,D., Professor of Early Childhood, Elementary and Special Education, Lee University, wrote the Asset Mapping Literature Review and contributed to survey development. For Contributions to Survey Development: Anne-Christine Bataillard, Lausanne Senior Associate for Children, President of Grain of Wheat Int l Mike Iosia, Ph.D., CSCS, Associate Professor of Exercise Physiology, Lee University, Cleveland, TN Miles Kirby, M.S., Research Assistant, London School of Hygiene and Tropical Medicine Kevin E Lawson, Professor of Christian Education, Director, Ph.D and Ed.D. programs in Educational Studies Richard Luna, Regional Director for Latin America, OneHope Kimberly McKay, M.Ed., Research Project Manager, OneHope Chris Odell, M.P.P, independent consultant: Research and Evaluation Design. Colin Piper, Chairman, World Evangelical Alliance Youth Commission Menchit Wong, International Advocacy Director for Compassion International 60

Global Study of Ministry to Children and Youth Oversight Chad Causey, Vice President for Global Ministry, OneHope Tyler Prieb, Director for International Partnership, OneHope Christopher Bedzyk, Project Portfolio Manager, Finished Consulting Services, LLC Dave Plate, M.P.H, Research and Evaluation Coordinator, OneHope. 61

APPENDIX A OneHope s Definition of Missional Reach For ministry Reach to be considered missional it must meet the following criteria: Redemptive story of Christ with tested cognitive increase (at design stage) Designed for ages 5 19 Outcome designed and evaluated Plus four of the following six: Opportunity for life transformation (salvation plan) Presentation of life transformation (testimony) Opportunity for integration into community Ongoing engagement through incremental activities Including one of the following: Printed product with: o 1,000 words of Scripture for lower primary o 2,000 words of Scripture for upper primary o 3,000 words of Scripture for teens Robust Scripture content 62

APPENDIX B Colombia Asset Mapping Survey Survey of Ministry to Children & Youth in Colombia 63

64

65

66