Spirit and Power: A 10-Country Survey of Pentecostals

Data Archive > International Surveys and Data > Multiple Nation Surveys > Summary


This multi-country survey was commissioned by the Pew Forum on Religion and Public Life to investigate the religious, political, and civic views of renewalists (i.e., Pentecostals and Charismatics) around the world. The project includes surveys in ten countries with sizeable renewalist populations: the United States; Brazil, Chile and Guatemala in Latin America; Kenya, Nigeria and South Africa in Africa; and India, the Philippines and South Korea in Asia. In each country, surveys were conducted among a random sample of the general public, with an oversample of renewalists to yield sufficient sample sizes for analysis. Surveys were conducted under the direction of Princeton Survey Research Associates International. The codebook reflects the results of the general public sample in each country.

Data File
Cases: 11,028
Variables: 200
Weight Variable: GPWGT and PENTWGT
Weight notes: GPWGT is the weight variable for use in analyzing general public samples in each country. PENTWGT is the weight variable for use in analyzing renewalists in each country.
Data Collection
Date Collected: May 5 – September 7, 2006
Funded By
Produced by the Pew Forum on Religion and Public Life with support from The Pew Charitable Trusts and The John Templeton Foundation.
Sampling Procedures
“Results for the surveys are based on telephone interviews in the United States and face-to-face interviews in all other nations, conducted under the direction of Princeton Survey Research Associates International. The surveys in Guatemala, Nigeria, the Philippines and the United States are based on national samples. The survey in Chile is based on a national sample but excludes non-continental and remote areas, and the survey in Kenya is based on a national sample but excludes the largely Muslim North Eastern Province. In Brazil, South Africa and South Korea, the surveys are based on urban samples.

“In India, the survey was conducted in three states believed to have among the highest percentage of Christians in India: Tamil Nadu, Kerala and Meghalaya. Within the three selected states, districts with the highest proportion of Christians were first selected, and then sampling points were randomly selected from these districts. This survey is NOT representative of the general population of India, nor is it representative of the population of the three Indian states in which it was conducted.” (See below for country-specific details. Quotations are from methodology reports provided by Pew Forum on Religion and Public Life.)
Principal Investigators
Pew Forum on Religion and Public Life
Related Publications
The Pew Forum on Religion and Public Life Report: Spirit and Power: A 10-Country Survey of Pentecostals, October 2007 http://pewforum.org/newassets/surveys/pentecostal/pentecostals-08.pdf

Complete summary information is available at the website:
http://pewforum.org/surveys/pentecostal/

Methodology information is available at the website:
http://www.pewforum.org/Christian/Evangelical-Protestant-Churches/Spirit-and-Power-A-10-Country-Survey-of-Pentecostals(4).aspx
SURVEY DETAILS BY COUNTRY
(All text below provided by Pew Forum on Religion and Public Life.)

Brazil
Survey firm: Research International Brazil
Sample design: Probability sample of urban cities
Mode: Face-to-face with adults 18+
Language: Portuguese
Fieldwork dates: May 13 – May 30, 2006
Data processing: June 1 – July 25, 2006
Sample size: General public – 700; Pentecostals – 313 (106 from the general public sample and 207 from the oversample of renewalists); Charismatics – 329 (236 from the general public sample and 93 from the oversample of renewalists)
Margin of error: General public 4%, Pentecostals 6%, Charismatics 5%

Chile
Survey firm: MORI Chile
Sample design: National probability sample, excluding non-continental and remote areas
Mode: Face-to-face with adults 18+
Language: Spanish
Fieldwork dates: May 12 – May 28, 2006
Data processing: June 28 – July 25, 2006
Sample size: General public – 600; Pentecostals – 276 (51 from the general public sample and 225 from the oversample of renewalists); Charismatics – 286 (123 from the general public sample and 163 from the oversample of renewalists)
Margin of error: General public 4%, Pentecostals 6%, Charismatics 6%

Guatemala
Survey firm: MERCAPLAN
Sample design: National probability sample
Mode: Face-to-face with adults 18+
Language: Spanish
Fieldwork dates: May 5 – May 31, 2006
Data processing: June 26 – July 25, 2006
Sample size: General public – 1,005; Pentecostals – 410 (198 from the general public sample and 212 from the oversample of renewalists); Charismatics – 487 (399 from the general public sample and 88 from the oversample of renewalists)
Margin of error: General public 3%, Pentecostals 5%, Charismatics 4%

India
Survey firm: TNS India
Sample design: Probability sample of disproportionately Christian districts of three states—Tamil Nadu, Kerala and Meghalaya
Mode: Face to face with adults 18+
Languages: Hindi, Tamil and Malayalam
Fieldwork dates: May 16 – June 9, 2006
Data processing: July 5 – July 25, 2006
Sample size: General public – 726; Pentecostals – 409 (10 from the general public sample and 399 from the oversample of renewalists); Charismatics – 125 (26 from the general public sample and 99 from the oversample of renewalists)
Margin of error: General public 4%, Pentecostals 5%, Charismatics 9%

Kenya
Survey firm: Research Path Associates
Sample design: National probability sample, excluding North Eastern Province
Mode: Face-to-face with adults 18+
Languages: English and Kiswahili
Fieldwork dates: May 17 – May 26, 2006
Data processing: June 29 – July 25, 2006
Sample size: General public – 655; Pentecostals – 403 (208 from the general public sample and 195 from the oversample of renewalists); Charismatics – 306 (155 from the general public sample and 151 from the oversample of renewalists)
Margin of error: General public 4%, Pentecostals 5%, Charismatics 6%

Nigeria
Survey firm: Research International Nigeria
Sample design: National probability sample
Mode: Face-to-face with adults 18+
Languages: Yoruba, Igbo, Hausa and English
Fieldwork dates: May 15 – July 27, 2006
Data processing: July 31 – August 4, 2006
Sample size: General public – 650; Pentecostals – 483 (139 from the general public sample and 334 from the oversample of renewalists); Charismatics – 67 (62 from the general public sample and 5 from the oversample of renewalists)
Margin of error: General public 4%, Pentecostals 4%, Charismatics N/A (Due to the small number of Nigerian charismatics surveyed, results for Nigerian charismatics are not presented in Pew Forum reporting on this survey.)

Philippines
Survey firm: TNS Philippines
Sample design: National probability sample
Mode: Face-to-face with adults 18+
Languages: Tagalog, Cebuano and Ilonggo
Fieldwork dates: May 6 – May 29, 2006
Data processing: June 19 – July 25, 2006
Sample size: General public – 1,000; Pentecostals – 309 (38 from the general public sample and 271 from the oversample of renewalists); Charismatics – 433 (404 from the general public sample and 29 from the oversample of renewalists)
Margin of error: General public 3%, Pentecostals 6%, Charismatics 5%

South Africa
Survey firm: Markinor
Sample design: National probability sample of urban areas
Mode: Face-to-face with adults 18+
Languages: Afrikaans, Pedi, Sotho, Tswana, Xhosa, Zulu and English
Fieldwork dates: May 11 – May 27, 2006
Data processing: June 26 – July 25, 2006
Sample size: General public – 800; Pentecostals – 259 (84 from the general public sample and 175 from the oversample of renewalists); Charismatics – 344 (182 from the general public sample and 162 from the oversample of renewalists)
Margin of error: General public 3%, Pentecostals 6%, Charismatics 5%

South Korea
Survey firm: Gallup Korea
Sample design: National probability sample of urban cities
Mode: Face-to-face with adults 18+
Language: Korean
Fieldwork dates: May 8 – May 26, 2006
Data processing: June 18 – July 25, 2006
Sample size: General public – 600; Pentecostals – 131 (13 from the general public sample and 118 from the oversample of renewalists); Charismatics – 333 (51 from the general public sample and 282 from the oversample of renewalists)
Margin of error: General public 4%, Pentecostals 9%, Charismatics 5%

United States
Survey firm: Princeton Survey Research Associates International
Sample design: National probability sample
Mode: Telephone
Languages: English and Spanish
Fieldwork dates: July 20 – September 7, 2006
Data processing: September 8 – September 12, 2006
Sample size: General public – 739; Pentecostals – 119 (41 from the general public sample and 78 from the oversample of renewalists); Charismatics – 421 (164 from the general public sample and 257 from the oversample of renewalists)
Margin of error: General public 4%, Pentecostals 9%, Charismatics 5%

Notes on Analysis and Weighting

(All text below provided by Pew Forum on Religion and Pubic Life.)

Since this dataset combines results from surveys conducted among the public at large in each country with results from surveys conducted among oversamples of renewalists in each country, users must exercise caution when analyzing the data to ensure that they have selected for the desired sample. To analyze results from the general public samples, users should select for those cases where the variable “sample” is equal to 1. To analyze results for renewalists, users should select for those cases where the variable “renewalist” is greater than 0.

Where necessary, survey results were weighted to match the demographic characteristics of the population in each country. The dataset contains two weight variables, GPWGT and PENTWGT. Analysis that focuses solely on the renewalist respondents should use PENTWGT, which provides a non-zero weight for all renewalist respondents in both the general public and oversamples. Non-renewalist respondents have a weight of zero on PENTWGT.

Analysis that focuses on the general public in each of the countries, irrespective of whether a given respondent is a renewalist or not, should use GPWGT. This weight allows for estimates of the proportion of renewalists in the general public or in subgroups of the general public. The renewalists interviewed as part of the overamples have a weight of zero on GPWGT and thus are excluded from this type of analysis using the general public sample.

There is no weight variable that combines the renewalist oversamples with the full general population samples in each country, nor is there a weight variable that combines the ten countries into a single international sample.