Sangeeta Agrawal, M.S., Gallup James K. Harter, Ph.D., Gallup

Transcription

1 Wellbeing Meta-Analysis: A Worldwide Study of the Relationship Between the Five Elements of Wellbeing and Life Evaluation, Daily Experiences, Health, and Giving Sangeeta Agrawal, M.S., Gallup James K. Harter, Ph.D., Gallup April 2011

2 Through advanced social and economic analysis, Gallup helps organizations, cities, and countries solve the world s foremost problems. For more information, please visit socialandeconomicanalysis.gallup.com or contact Sarah Van Allen at or sarah_van_allen@gallup.com. Copyright Standards This document contains proprietary research, copyrighted materials, and literary property of Gallup, Inc. It is for the guidance of your company only and is not to be copied, quoted, published, or divulged to others outside of your organization. Gallup, Wellbeing Finder, Gallup Panel, and Gallup Consulting are trademarks of Gallup, Inc. All other trademarks are the property of their respective owners. This document is of great value to both your organization and Gallup, Inc. Accordingly, international and domestic laws and penalties guaranteeing patent, copyright, trademark, and trade secret protection protect the ideas, concepts, and recommendations related within this document. No changes may be made to this document without the express written permission of Gallup, Inc.

3 Table of Contents Executive Summary... 1 Objective...1 Methods...1 Results...1 Conclusion...2 Introduction... 2 Wellbeing Finder...2 World Poll Methodology...2 Preparing for Data Collection...3 Translation...3 Data Collection Methodology, Study Sample, and Database...3 Meta-Analysis Variables...3 Outcome-Dependent Variables...3 Predictor Variables...5 Demographic Variables...5 Meta-Analysis, Hypothesis, Methods, and Results... 5 Meta-Analysis...5 Hypothesis and Study Characteristics...6 Meta-Analytic Methods Used...7 Reliability Calculations...7 Results...7 Relation With Country-Level Moderator Variable...13 GDP Variable...13 References...18 Appendix A: Sample Size by Country or Region...19 Appendix B: The Gallup Panel...21 Copyright 2011 Gallup, Inc. All rights reserved.

4 Copyright 2011 Gallup, Inc. All rights reserved.

5 Wellbeing Meta-Analysis: A Worldwide Study of the Relationship Between the Five Elements of Wellbeing and Life Evaluation, Daily Experiences, Health, and Giving Sangeeta Agrawal, M.S., Gallup James K. Harter, Ph.D., Gallup Executive Summary Objective Wellbeing is all the things that are important to how we think about and experience our lives. Previous Gallup research found five actionable elements that differentiate thriving lives from those spent struggling and suffering. This report provides an international cross-validation of the five elements of wellbeing. The purpose of this study was to: 1. estimate the relationship between each of the five elements of wellbeing Career, Social, Financial, Physical, and Community and six outcome variables present life evaluation, future life evaluation, daily wellbeing, unhealthy days, health problems, and giving across countries and regions worldwide 2. estimate the practical meaning of the relationship between the wellbeing elements and the outcome variables 3. examine the moderating effect of country-level GDP and location/region on the relationship between each of the wellbeing elements and the outcome variables Methods The data set includes nationally representative samples of adults, aged 15 and older, in 117 countries from the Gallup World Poll, representing more than 95% of the world s adult population. Gallup conducted 120,239 interviews from February 2009 through March 2010 using face-to-face or telephone methodology. The Gallup World Poll survey includes a core set of approximately 100 wellbeing items. In 2009, based on research that found five important subjective factors, or elements, of wellbeing, Gallup added five items to the World Poll survey one item that best explained variance in each of the five elements of wellbeing. We statistically calculated the individual-level relationship between the five wellbeing items and the six outcome variables, including present life evaluation, future life evaluation, daily wellbeing, number of unhealthy days in the past 30 days, general health problems, and giving. We examined the moderation of the previously mentioned relationships across country-level GDP and location/region categorization. Results Each of the five wellbeing elements explains meaningful variance in each of the six outcome variables studied. Results indicate high generalizability, indicating that the correlations were largely consistent across different countries and areas after controlling for demographic differences. Respondents with higher wellbeing in all five areas reported higher present life and future life evaluations, better daily experiences, fewer unhealthy days and health problems, and higher likelihood of giving. Those with high Career Wellbeing Copyright 2011 Gallup, Inc. All rights reserved. 1

6 reported less than half as many unhealthy days and were 39% less likely to report chronic health problems than those with low Career Wellbeing. Those with high Social Wellbeing were 17% more likely to report giving than those with low Social Wellbeing. Those with high Financial Wellbeing reported higher life satisfaction than those with low Financial Wellbeing, even after controlling for income and other demographics and regional differences. Conclusion The relationship between the five wellbeing elements and the six outcome variables at the individual level is substantial and generalizable across countries and areas. This relationship provides further evidence of the universality of the five core elements of wellbeing. Introduction The purpose of this study was to further cross-validate the five elements of wellbeing Career, Social, Financial, Physical, and Community in relation to six outcome variables present life evaluation, future life evaluation, daily wellbeing, unhealthy days, health problems, and giving. This study integrates two modes of Gallup wellbeing research: one focused on in-depth measurement for individuals Gallup Wellbeing Finder and the other focused on broad population reporting of wellbeing Gallup World Poll. To capture the opinions of a wide spectrum of the global population, we used the Gallup World Poll as the measurement vehicle. We used the Wellbeing Finder research to identify five core World Poll items to efficiently measure the five elements of wellbeing across a broad worldwide population. Wellbeing Finder The goal in developing the Gallup Wellbeing Finder was to leverage current behavioral science to build a comprehensive, reliable, valid, concise, and actionable tool that individuals could use to track their own wellbeing over time. Gallup s team of scientists sought to identify wellbeing dimensions that explain differences in wellbeing for people in many different life situations and that represent actionable areas that individuals can implement for wellbeing improvement. Wellbeing is all the things that are important to how we think about and experience our lives. Development of the Gallup Wellbeing Finder occurred in three iterations or phases: Phase 1: Review of historical Gallup wellbeing research Phase 2: Gallup World Poll analysis Phase 3: Pilot research for the Web assessment Pilot 1: Item testing across diverse groups Pilot 2: Refinement of measures/constructs in U.S. and international samples World Poll Methodology The Gallup World Poll continually surveys residents in more than 150 countries, representing approximately 98% of the world s adult population, using randomly selected, nationally representative samples. Gallup typically surveys 1,000 2 Copyright 2011 Gallup, Inc. All rights reserved.

7 individuals in each country using a standard set of core questions that has been translated into the major languages of the respective country. In some regions, supplemental questions are asked in addition to core questions. Face-to-face interviews are approximately 1 hour, while telephone interviews are about 30 minutes. In many countries, the survey is conducted once per year, and fieldwork is generally completed in two to four weeks. Gallup is responsible for the management, design, and control of the Gallup World Poll. For the past 70 years, Gallup has been committed to the principle that accurately collecting and disseminating the opinions and aspirations of people around the globe is vital to understanding our world. Gallup s mission is to provide information in an objective, reliable, and scientifically grounded manner. Gallup is not associated with any political orientation, party, or advocacy group and does not accept partisan entities as clients. Any individual, institution, or governmental agency can access the Gallup World Poll regardless of nationality. The identities of clients and all surveyed respondents are confidential. Preparing for Data Collection Translation The questionnaire is translated into the major languages of each country. The translation process starts with an English, French, or Spanish version, depending on the region. A translator who is proficient in the original and target languages translates the survey into the target language. A second translator reviews the language version against the original version and recommends refinements. The Worldwide Research Methodology and Codebook provides additional methodological detail for the World Poll. Data Collection Methodology, Study Sample, and Database The data set includes nationally representative samples of adults, aged 15 and older, in 117 countries from the Gallup World Poll, representing more than 95% of the world s adult population. Gallup conducted 120,239 interviews from February 2009 through March 2010 using face-to-face or telephone methodology. Meta-Analysis Variables Outcome-Dependent Variables There are six outcome-dependent variables in this analysis: 1. Present Life Evaluation (ladder scale) Please imagine a ladder, with steps numbered from 0 at the bottom to 10 at the top. The top of the ladder represents the best possible life for you and the bottom of the ladder represents the worst possible life for you. On which step of the ladder would you say you personally feel you stand at this time? Copyright 2011 Gallup, Inc. All rights reserved. 3

8 2. Future Life Evaluation (ladder scale) Please imagine a ladder, with steps numbered from 0 at the bottom to 10 at the top. The top of the ladder represents the best possible life for you and the bottom of the ladder represents the worst possible life for you. Just your best guess, on which step do you think you will stand in the future, say about five years from now? 3. Daily Wellbeing We obtained results on daily experiences based on 10 daily items (scored 0-10, where 1 is scored for each positive experience or absence of a negative experience and 0 is scored for each negative experience or absence of a positive experience), which when combined make up the Positive Experience Index and the Negative Experience Index to form a composite of experiences yesterday. Daily Individual Items: Positive Experience Index Now, please think about yesterday, from the morning until the end of the day. Think about where you were, what you were doing, who you were with, and how you felt. A. Did you feel well-rested yesterday? B. Were you treated with respect all day yesterday? C. Did you smile or laugh a lot yesterday? D. Did you learn or do something interesting yesterday? E. Did you experience the following feelings a lot of the day yesterday? How about enjoyment? 1 Yes 2 No Daily Individual Items: Negative Experience Index Did you experience the following feelings during a lot of the day yesterday? How about... A. Physical pain B. Worry C. Sadness D. Stress E. Anger 1 Yes 2 No 4 Copyright 2011 Gallup, Inc. All rights reserved.

9 4. Number of Unhealthy Days During the past 30 days, for about how many days did poor health keep you from doing your usual activities? 5. Health Problems Do you have any health problems that prevent you from doing any of the things people your age normally can do? 6. Giving Have you done any of the following in the past month? How about...? A. Donated money to a charity B. Volunteered your time to an organization C. Helped a stranger or someone you didn t know who needed help 1 Yes 2 No Predictor Variables There are five predictor variables each representing one wellbeing element. Thinking about your life in general, please rate your level of agreement with each of the following using a five point scale, where five means you STRONGLY AGREE and one means you STRONGLY DISAGREE. A. I like what I do each day. (Career) B. I have a lot of love in my life. (Social) C. I have more than enough money to do what I want to do. (Financial) D. My physical health is near perfect. (Physical) E. I can t imagine living in a better community than the one I live in today. (Community) Demographic Variables Demographic variables include age, gender, marital status, education, and income level. Meta-Analysis, Hypothesis, Methods, and Results Meta-Analysis A meta-analysis is a statistical integration of data accumulated across numerous studies from countries where Gallup surveys. As such, it provides uniquely powerful information because it controls for measurement and sampling errors Copyright 2011 Gallup, Inc. All rights reserved. 5

10 and other idiosyncrasies that distort the results of individual studies. A meta-analysis eliminates biases and provides an estimate of true validity or true relationship between two or more variables. Statistics typically calculated during metaanalyses also allow researchers to explore the presence or lack of moderators of relationships. More than 1,000 meta-analyses have been conducted in the psychological, educational, behavioral, medical, and personnel selection fields. The research literature in the behavioral and social sciences fields includes a multitude of individual studies with apparently conflicting conclusions. Meta-analysis, however, allows researchers to estimate the mean relationship between variables and make corrections for artifactual sources of variation in findings across studies. It provides a method by which researchers can determine whether validities and relationships generalize across various situations (e.g., countries or geographical locations). Individual studies often contain different sample sizes and idiosyncrasies that distort the interpretation of results. Metaanalysis is a statistical technique that is useful in combining results of studies with seemingly disparate findings, correcting for sampling, measurement error, and other study artifacts to understand the true relationship with greater precision. We applied Hunter-Schmidt meta-analysis methods to partial correlations calculated from 117 countries and areas to estimate the true relationship between the five wellbeing elements and outcome variables and to test for generalizability after controlling for demographic differences. Hypothesis and Study Characteristics We examined the following hypotheses for this meta-analysis: Hypothesis 1: The five wellbeing elements will have positive correlations with the following outcome variables: present life evaluation, future life evaluation, daily wellbeing, and giving. The five wellbeing elements will have significant negative correlations with the following outcome variables: number of unhealthy days, and health problems. Hypothesis 2: The correlations between the five wellbeing elements and the six outcome variables will generalize across countries. That is, these correlations will not vary substantially across countries. And, in particular, there will be few, if any, countries with zero correlations or those in the opposite direction from Hypothesis 1. In total, 117 countries including 120,239 respondents were included in the meta-analysis. Appendix A includes the sample size for each country. We calculated partial correlations, estimating the relationship of each of the predictor variables to each of these six outcomes after controlling for demographic differences Table 1 Sample Characteristics of Variables in the Model About the database n Mean SD Demographic variables 120,239 Numbers of respondents/country 117 1, Variables Outcome Variables Present life evaluation 118, Future life evaluation 109, Daily wellbeing 117, Number of unhealthy days 94, Health problems (yes)* 119, Giving (yes)* 117, Predictor Variables Career Wellbeing 119, Social Wellbeing 119, Financial Wellbeing 119, Physical Wellbeing 119, Community Wellbeing 117, Demographic Variables Age 120, Gender (male)* 120, Marital status (married)* 119, Education (high school or more)* 119, Income (annual median household income in USD) $119,239 $7, *Means represent proportion of respondents 6 Copyright 2011 Gallup, Inc. All rights reserved.

11 within each country. We entered these correlation coefficients into a database and then calculated mean validities, standard deviations of validities, and validity generalization statistics for each of the correlations. Meta-Analytic Methods Used Analyses included mean observed and estimates of true score correlation, estimates of standard deviation of validities, and corrections made for sampling error and measurement error in the dependent variables. We conducted an additional analysis to correct for independent-variable measurement error. The most basic form of meta-analysis corrects variance estimates only for sampling error. Other corrections recommended by Hunter and Schmidt include correction for measurement and statistical artifacts such as range restriction and measurement error in the dependent variables gathered. We used the Schmidt and Le meta-analysis package for our meta-analysis. Reliability Calculations We used two Gallup Panel surveys to calculate test-retest reliability estimates for all of the five elements of wellbeing, or predictor variables, and four of the six outcome variables (Table 2). Gallup Panel participants completed two independent surveys administered in January and February 2009: a wellbeing survey (n=2,307) and a health survey (n=45,980). We matched responses for those participants who completed both surveys to calculate test-retest reliability. The average number of days difference between the two surveys was 21 days (median=20; range=8-54), therefore these estimates of test-retest reliability allow for correction for transient error in addition to random response error because the time period between test and retest is less than one would expect real change to occur on the variables measured. Table 2 Reliability Estimates Variables Wellbeing Survey Health Survey Reliability Estimate Mean SD Mean SD n Correlation Outcome Variables Present life evaluation , Future life evaluation , Number of unhealthy days Health problems (yes)* % % 1, * Predictor Variables Career Wellbeing , Social Wellbeing** , Financial Item , Physical Wellbeing , Community Wellbeing*** , *Chronbach s alpha ** Health survey question was different: The relationships in my life give me positive energy every day. *** Health survey question was different: I can t imagine a stronger community than the one I live in today. Results The focus of analyses for this report is on the relationship between the five predictor variables (wellbeing elements) and various outcome variables. Table 3 provides meta-analytic and validity generalization statistics for the relationship between the predictor variables and each of the six outcome variables studied. In each of the tables that follow, we provide various meta-analytic statistics. The statistics provided are as follows: Number of Respondents: number of respondents to the survey Number of (Countries) r s: number of correlations (countries) studied for each item Mean Observed r: the sample size weighted correlation of each item to the dependent variable uncorrected for measurement error Copyright 2011 Gallup, Inc. All rights reserved. 7

12 Observed SD (standard deviation): the sample size weighted standard deviation of observed correlations corrected for the sample size of each study True Score Correlation: the sample size weighted correlation of each item to the dependent variable after correcting for dependent variable measurement error and independent variable measurement error SD of True Score Correlation: the sample size weighted standard deviation of the correlations, correcting for dependent variable measurement error and independent variable measurement error % Variance Acc d for: the percentage of variance in correlations across countries accounted for by sampling error across countries 90% CV: the 90% credibility value (10 th percentile of the distribution of true validities) Table 3 Meta-Analysis and Validity Generalization Statistics: Relationship Between Five Wellbeing Elements and Present Life Evaluation (partial correlations controlling for demographics) Outcome Variable Career Social Financial Physical Community Number of Respondents 109, , , , ,455 Number of (Countries) r's Mean Observed r Observed SD True Score Correlation SD of True Score Correlation % Variance Acc'd for % CV Key Finding: Each of the five wellbeing elements had evidence of generalizability in relationship to respondents overall present life evaluations. Financial and Career Wellbeing had the strongest true score correlations with overall life evaluations. Table 4 Meta-Analysis and Validity Generalization Statistics: Relationship Between Five Wellbeing Elements and Future Life Evaluation (partial correlations controlling for demographics) Outcome Variable Career Social Financial Physical Community Number of Respondents 105, , , , ,408 Number of (Countries) r's Mean Observed r Observed SD True Score Correlation SD of True Score Correlation % Variance Acc'd for % CV Key Finding: Each of the five wellbeing elements had evidence of generalizability in relationship to respondents life evaluation expectations for the next five years. True score correlations were strongest for Social, Financial, and Physical Wellbeing. 8 Copyright 2011 Gallup, Inc. All rights reserved.

13 Table 5 Meta-Analysis and Validity Generalization Statistics: Relationship Between Five Wellbeing Elements and Daily Experiences (partial correlations controlling for demographics) Outcome Variable Career Social Financial Physical Community Number of Respondents 108, , , , ,387 Number of (Countries) r's Mean Observed r Observed SD True Score Correlation SD of True Score Correlation % Variance Acc'd for % CV Key Finding: Each of the five wellbeing elements had evidence of generalizability in relationship to respondents daily experiences. True score correlations were strongest for Career, Physical, and Social Wellbeing. Table 6 Meta-Analysis and Validity Generalization Statistics: Relationship Between Five Wellbeing Elements and Unhealthy Days (partial correlations controlling for demographics) Outcome Variable Career Social Financial Physical Community Number of Respondents 86,431 86,171 86,373 86,455 90,798 Number of (Countries) r's Mean Observed r Observed SD True Score Correlation SD of True Score Correlation % Variance Acc'd for % CV Key Finding: The true score relationship between each of the five wellbeing elements and unhealthy days was negative higher wellbeing is associated with fewer unhealthy days in the last 30 days. Three of the five wellbeing elements Physical, Career, and Financial had evidence of generalizability across countries in relationship to unhealthy days. True score correlations were strongest for Physical Wellbeing. Copyright 2011 Gallup, Inc. All rights reserved. 9

14 Table 7 Meta-Analysis and Validity Generalization Statistics: Relationship Between Five Wellbeing Elements and Health Problems (partial correlations controlling for demographics) Outcome Variable Career Social Financial Physical Community Number of Respondents 109, , , , ,524 Number of (Countries) r's Mean Observed r Observed SD True Score Correlation SD of True Score Correlation % Variance Acc'd for % CV Key Finding: The true score relationship between each of the five wellbeing elements and respondents selfreported health problems was negative higher wellbeing is associated with fewer health problems. Four of the five wellbeing elements Physical, Career, Financial, and Social had evidence of generalizability across countries in relationship to health problems. True score correlations were strongest for Physical Wellbeing. Table 8 Meta-Analysis and Validity Generalization Statistics: Relationship Between Five Wellbeing Elements and Giving (partial correlations controlling for demographics) Outcome Variable Career Social Financial Physical Community Number of Respondents 108, , , , ,313 Number of (Countries) r's Mean Observed r Observed SD True Score Correlation SD of True Score Correlation % Variance Acc'd for % CV Key Finding: The relationship between each of the five wellbeing elements and respondents self-reported giving in the last 30 days was positive higher wellbeing is associated with more frequent reports of giving. Two of the five wellbeing elements Career and Social had evidence of generalizability across countries in relationship to giving. True score correlations were strongest for Career and Social Wellbeing. 10 Copyright 2011 Gallup, Inc. All rights reserved.

15 Table 9 Meta-Analytic True Score Partial Correlations With Each Element After Controlling for Demographic Differences Outcome Variable Independent Variables Career Social Financial Physical Community Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Key Finding: This table summarizes the true score partial correlations for each wellbeing element and outcome variable combination. While the magnitude of the correlations varies after controlling for demographic variables, all of the true score correlations are in the hypothesized direction. Practical Meaning of Effects To understand the practical meaning of the meta-analytic partial correlations, we plotted the dependent variable averages across the scale continuum for each wellbeing element. Table 10 Outcome Variables for Like What I Do Each Day Outcome variables (controlling for demographic and regional differences) 1.00 Strongly disagree Strongly agree Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) 37.0% 34.2% 28.4% 23.3% 22.5% Giving (yes) 54.6% 50.1% 53.4% 56.2% 64.1% Key Finding: The difference between those with high Career Wellbeing and those with low Career Wellbeing is 1.5 percentage points on the 0-10 life evaluation scale, 2.2 points for daily experiences on a 0-10 scale, 3.2 unhealthy days in the last 30 days, 14.5 points in reported health problems, and 9.5 points in reported giving. Copyright 2011 Gallup, Inc. All rights reserved. 11

16 Table 11 Outcome Variables for Lot of Love in My Life Outcome variables (controlling for demographic and regional differences) 1.00 Strongly disagree Strongly agree Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) 33.6% 32.0% 27.7% 24.6% 23.8% Giving (yes) 54.4% 49.4% 52.0% 55.7% 63.4% Key Finding: The difference between those with high Social Wellbeing and those with low Social Wellbeing is 1.0 points on the 0-10 life evaluation scale, 1.7 points for daily experiences on a 0-10 scale, 2.1 unhealthy days in the last 30 days, 9.8 points in reported health problems, and 9 points in reported giving. Table 12 Outcome Variables for Have More Than Enough Money Outcome variables (controlling for demographic and regional differences) 1.00 Strongly disagree Strongly agree Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) 31.0% 25.2% 23.2% 21.7% 22.8% Giving (yes) 53.3% 55.7% 60.5% 62.9% 64.5% Key Finding: The difference between those with high Financial Wellbeing and those with low Financial Wellbeing is 1.8 points on the 0-10 life evaluation scale, 1.6 points for daily experiences on a 0-10 scale, 1.5 unhealthy days in the last 30 days, 8.2 points in reported health problems, and 11.2 points in reported giving. Table 13 Outcome Variables for Physical Health Near Perfect Outcome variables (controlling for demographic and regional differences) 1.00 Strongly disagree Strongly agree Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) 63.6% 53.4% 30.7% 15.2% 13.4% Giving (yes) 52.4% 53.7% 56.1% 57.8% 62.4% Key Finding: The difference between those with high Physical Wellbeing and those with low Physical Wellbeing is 1.2 points on the 0-10 life evaluation scale, 2.4 points for daily experiences on a 0-10 scale, 7.8 unhealthy days in the last 30 days, 50.2 points in reported health problems, and 10 points in reported giving. 12 Copyright 2011 Gallup, Inc. All rights reserved.

17 Table 14 Outcome Variables for Can t Imagine a Better Community Outcome variables (controlling for demographic and regional differences) 1.00 Strongly disagree Strongly agree Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) 28.0% 27.3% 26.1% 24.2% 23.8% Giving (yes) 58.9% 53.9% 54.8% 57.5% 63.7% Key Finding: The difference between those with high Community Wellbeing and those with low Community Wellbeing is 0.8 points on the 0-10 life evaluation scale, 1.0 points for daily experiences on a 0-10 scale, 0.8 unhealthy days in the last 30 days, 4.2 points in reported health problems, and 4.8 points in reported giving. Relation With Country-Level Moderator Variable We examined the moderating effect of the GDP variable to understand the variation in relationship between the five wellbeing elements and the outcome variables. This analysis includes 104 countries because GDP values were not available for 13 of the 117 countries surveyed. GDP Variable According to the World Bank, purchasing power parity (PPP) GDP is gross domestic product converted to international dollars using PPP rates. An international dollar has the same purchasing power over GDP as the U.S. dollar has in the United States. GDP at purchaser s prices is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources. First, we calculated partial correlations, estimating the relationship between each of the five wellbeing elements and the outcome variables within each country. Next, we correlated country-level partial correlations and GDP. Table 15 shows the correlation between GDP and the partial correlations across countries. For instance, the correlations between Career Wellbeing and present life evaluation were higher in high GDP countries (r=.378) even though the correlation between Career Wellbeing and present life evaluation was generalizable in the positive direction across countries. Copyright 2011 Gallup, Inc. All rights reserved. 13

18 Table 15 Correlations Between GDP and Partial Correlations Across Countries and Areas Correlations Dependent Variables Correlation With GDP Career Wellbeing Partial correlation between present life evaluation and Career Wellbeing.378** Partial correlation between future life evaluation and Career Wellbeing.277** Partial correlation between daily wellbeing and Career Wellbeing.210* Partial correlation between "days poor health kept from activity" and Career Wellbeing.173 Partial correlation between health problems and Career Wellbeing Partial correlation between giving/donated money/helped a stranger/volunteered time and Career Wellbeing -.195* Social Wellbeing Partial correlation between present life evaluation and Social Wellbeing.399** Partial correlation between future life evaluation and Social Wellbeing.344** Partial correlation between daily wellbeing and Social Wellbeing.121 Partial correlation between "days poor health kept from activity" and Social Wellbeing.033 Partial correlation between health problems and Social Wellbeing Partial correlation between giving/donated money/helped a stranger/volunteered time and Social Wellbeing Financial Wellbeing Partial correlation between present life evaluation and Financial Wellbeing.448** Partial correlation between future life evaluation and Financial Wellbeing.298** Partial correlation between daily wellbeing and Financial Wellbeing.365** Partial correlation between "days poor health kept from activity" and Financial Wellbeing Partial correlation between health problems and Financial Wellbeing Partial correlation between giving/donated money/helped a stranger/volunteered time and Financial Wellbeing Physical Wellbeing Partial correlation between present life evaluation and Physical Wellbeing.424** Partial correlation between future life evaluation and Physical Wellbeing.335** Partial correlation between daily wellbeing and Physical Wellbeing.158 Partial correlation between "days poor health kept from activity" and Physical Wellbeing.299** Partial correlation between health problems and Physical Wellbeing Partial correlation between giving/donated money/helped a stranger/volunteered time and Physical Wellbeing Community Wellbeing Partial correlation between present life evaluation and Community Wellbeing.550** Partial correlation between future life evaluation and Community Wellbeing.422** Partial correlation between daily wellbeing and Community Wellbeing.329** Partial correlation between "days poor health kept from activity" and Community Wellbeing Partial correlation between health problems and Community Wellbeing -.327** Partial correlation between giving/donated money/helped a stranger/volunteered time and Community Wellbeing *p<0.01 **p<0.001 This moderation analysis indicates some moderation related to country-level GDP for 16 of the 30 partial correlations of the paired variables. The correlations are somewhat higher within higher GDP countries, while directionally generalizable. This is primarily the case for the life evaluation and daily experiences dependent variables, and was strongest across the dependent variables for the Community Wellbeing element. To better understand the moderation effect, we examined country-level variation in GDP values in Copyright 2011 Gallup, Inc. All rights reserved.

19 Graph 1 GDP Distribution Across Countries $40000 GDP per capita (in USD) $30000 $20000 $10000 $0 Countries with GDP <2,000 Countries with GDP 2,000-5,000 Countries with GDP 5,000-15,000 Countries with GDP 15,000+ We divided countries into four groups based on GDP values to illustrate the variation in correlation across countrylevel GDP groupings. Copyright 2011 Gallup, Inc. All rights reserved. 15

20 Table 16 Partial Correlation Between Each Outcome and Predictor Variable Combinations, by GDP Grouping Career Wellbeing Average partial correlation with Career Wellbeing after controlling for demographic differences GDP Group < 2,000 2,000-5,000 5,000-15,000 15,000+ Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Social Wellbeing Average partial correlation with Social Wellbeing after controlling for demographic differences GDP Group < 2,000 2,000-5,000 5,000-15,000 15,000+ Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Financial Wellbeing Average partial correlation with Financial Wellbeing after controlling for demographic differences GDP Group < 2,000 2,000-5,000 5,000-15,000 15,000+ Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Physical Wellbeing Average partial correlation with Physical Wellbeing after controlling for demographic differences GDP Group < 2,000 2,000-5,000 5,000-15,000 15,000+ Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Community Wellbeing Average partial correlation with Community Wellbeing after controlling for demographic differences GDP Group < 2,000 2,000-5,000 5,000-15,000 15,000+ Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Key Finding: Directionally, the partial correlations are similar within the different country-level GDP groupings, but correlations are somewhat different. Life evaluation correlations tend to be somewhat higher for higher GDP countries. Daily experiences and giving correlations are similar across GDP country groupings. Physical Wellbeing and health variable correlations are stronger in lower GDP countries 16 Copyright 2011 Gallup, Inc. All rights reserved.

21 Table 17 Correlations and Mean GDP by Region Average partial correlation with Career Wellbeing after controlling for demographic differences Sub-Saharan Africa Middle East and North Africa Japan and South Korea Career Wellbeing Southern Asia Southeast Asia Subregion China (includes Hong Kong) India Latin America Western Europe CIS* Central and Eastern Europe Mean GDP (in USD) $1,875.6 $10,051.8 $23,494.0 $2,362.1 $9,218.1 $18,541.8 $23,342.3 $6,473.8 $22,493.2 $4,405.1 $11,213.1 Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Average partial correlation with Social Wellbeing after controlling for demographic differences Sub-Saharan Africa Middle East and North Africa Japan and South Korea Social Wellbeing Southern Asia Southeast Asia Subregion China (includes Hong Kong) India Latin America Western Europe CIS* Central and Eastern Europe Mean GDP $1,875.6 $10,051.8 $23,494.0 $2,362.1 $9,218.1 $18,541.8 $23,342.3 $6,473.8 $22,493.2 $4,405.1 $11,213.1 Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Average partial correlation with Financial Wellbeing after controlling for demographic differences Sub-Saharan Africa Middle East and North Africa Japan and South Korea Financial Wellbeing Southern Asia Southeast Asia Subregion China (includes Hong Kong) India Latin America Western Europe CIS* Central and Eastern Europe Mean GDP $1,875.6 $10,051.8 $23,494.0 $2,362.1 $9,218.1 $18,541.8 $23,342.3 $6,473.8 $22,493.2 $4,405.1 $11,213.1 Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Average partial correlation with Physical Wellbeing after controlling for demographic differences Sub-Saharan Africa Middle East and North Africa Japan and South Korea Physical Wellbeing Southern Asia Southeast Asia Subregion China (includes Hong Kong) India Latin America Western Europe CIS* Central and Eastern Europe Mean GDP $1,875.6 $10,051.8 $23,494.0 $2,362.1 $9,218.1 $18,541.8 $23,342.3 $6,473.8 $22,493.2 $4,405.1 $11,213.1 Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) Average partial correlation with Community Wellbeing after controlling for demographic differences Sub-Saharan Africa Middle East and North Africa Japan and South Korea Community Wellbeing Southern Asia Southeast Asia Subregion China (includes Hong Kong) India Latin America Western Europe CIS* Central and Eastern Europe Mean GDP $1,875.6 $10,051.8 $23,494.0 $2,362.1 $9,218.1 $18,541.8 $23,342.3 $6,473.8 $22,493.2 $4,405.1 $11,213.1 Present life evaluation Future life evaluation Daily wellbeing Number of unhealthy days Health problems (yes) Giving (yes) *Commonwealth of Independent States and nearby countries Key Finding: While correlations vary somewhat by region, they are directionally similar across regions. This provides further evidence of the generalizability of the five wellbeing elements. Copyright 2011 Gallup, Inc. All rights reserved. 17

22 References Gallup. (2011, February). Worldwide research methodology and codebook. Omaha, NE: Author. Rath, T. & Harter, J. (2010). Wellbeing: The five essential elements. New York: Gallup Press. Hunter, J. E., & Schmidt, F. L. (1990). Methods of meta-analysis: Correcting error and bias in research findings. Newbury Park, CA: Sage. Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: Correcting error and bias in research findings (2nd ed.). Newbury Park, CA: Sage. Schmidt, F. L., & Le, H. A. (2004). Software for the Hunter-Schmidt meta-analysis methods. Iowa City, IA: Tippie College of Business, University of Iowa. The World Bank. (n.d.). GDP per capita, PPP (current international $). Retrieved March 15, 2011, from worldbank.org/indicator/ny.gdp.pcap.pp.cd?order=wbapi_data_value_2009+wbapi_data_value+wbapi_data_valuelast&sort=desc 18 Copyright 2011 Gallup, Inc. All rights reserved.

23 Appendix A: Sample Sizes by Country or Region Table 18 Sample Size by Country or Region Country or Region Sample Size Country or Region Sample Size Country or Region Sample Size United States 509 Niger 994 Libya 996 Egypt 1,032 Rwanda 1,000 Lithuania 498 Morocco 1,031 Senegal 995 Macedonia 1,000 Lebanon 1,005 Zambia 1,000 Malaysia 1,011 Saudi Arabia 1,021 South Korea 1,000 Mongolia 1,000 Jordan 1,001 Afghanistan 999 Montenegro 996 Syria 1,018 Belarus 1,074 Nepal 989 Turkey 996 Georgia 996 Nicaragua 1,012 Pakistan 1,972 Kazakhstan 997 Panama 1,017 Indonesia 1,078 Kyrgyzstan 1,000 Paraguay 999 Bangladesh 1,000 Moldova 999 Peru 999 United Kingdom 1,002 Russia 2,033 Portugal 997 France 1,000 Ukraine 1,072 Serbia 998 Germany 1,002 Cameroon 998 Slovenia 500 Spain 1,005 Zimbabwe 1,000 Sudan 1,000 Italy 1,005 Costa Rica 1,000 Switzerland 1,002 Poland 999 Albania 996 Tajikistan 994 Czech Republic 1,075 Algeria 1,000 Tunisia 1,004 Sweden 1,000 Argentina 1,000 Turkmenistan 1,000 Greece 1,000 Armenia 993 United Arab Emirates 1,041 Denmark 1,000 Austria 1,000 Uruguay 996 Hong Kong 755 Azerbaijan 990 Uzbekistan 993 Singapore 1,005 Bahrain 1,074 Yemen 1,000 Japan 1,000 Bolivia 997 Kosovo 975 China 3,842 Bosnia and Herzegovina 998 Somaliland region 1,000 India 2,995 Bulgaria 995 Total 120,239 Venezuela 997 Burundi 999 Brazil 1,031 Chad 998 Mexico 998 Chile 1,009 Nigeria 998 Colombia 1,000 Kenya 1,000 Comoros 1,000 Tanzania 1,000 Congo (Kinshasa) 1,000 Israel 1,000 Croatia 1,009 Palestinian Territories 995 Cyprus 500 Ghana 998 Djibouti 1,000 Uganda 1,000 Dominican Republic 1,000 Malawi 1,000 Ecuador 1,000 South Africa 1,000 El Salvador 1,006 Canada 500 Estonia 608 Philippines 1,000 Guatemala 1,011 Sri Lanka 995 Honduras 1,002 Vietnam 1,007 Iraq 995 Thailand 1,019 Ireland 500 Cambodia 1,000 Ivory Coast 1,000 Mali 999 Kuwait 1,001 Mauritania 984 Latvia 515 Copyright 2011 Gallup, Inc. All rights reserved. 19

24 Table 19 Sample Size by Worldwide Regions Sub- Saharan Africa Middle East and North Africa Japan and South Korea South Asia Southeast Asia China (includes Hong Kong) India United States and Canada Latin America Western Europe CIS* Central and Eastern Europe Nigeria Egypt Japan Pakistan Indonesia Hong Kong India United States Venezuela United Kingdom Belarus Turkey Kenya Morocco South Korea Bangladesh Singapore China Canada Brazil France Georgia Poland Tanzania Lebanon Sri Lanka Philippines Mexico Germany Kazakhstan Czech Republic Ghana Saudi Arabia Afghanistan Vietnam Costa Rica Spain Kyrgyzstan Greece Uganda Jordan Nepal Thailand Argentina Italy Moldova Albania Malawi Syria Cambodia Bolivia Sweden Russia Bosnia and Herzegovina South Africa Israel Malaysia Chile Denmark Ukraine Bulgaria Mali Palestinian Territories Colombia Austria Armenia Croatia Mauritania Algeria Dominican Republic Ireland Azerbaijan Cyprus Niger Bahrain Ecuador Portugal Mongolia Estonia Rwanda Iraq El Salvador Slovenia Tajikistan Latvia Senegal Kuwait Guatemala Switzerland Turkmenistan Lithuania Zambia Libya Honduras Uzbekistan Macedonia Cameroon Sudan Nicaragua Montenegro Zimbabwe Tunisia Panama Serbia Burundi United Arab Emirates Paraguay Kosovo Chad Yemen Peru Comoros Uruguay Congo (Kinshasa) Djibouti Ivory Coast Somaliland region 21,963 17,214 2,000 5,955 7,120 4,597 2,995 1,009 18,074 11,013 14,141 14,158 *Commonwealth of Independent States and nearby countries 20 Copyright 2011 Gallup, Inc. All rights reserved.

25 Appendix B: The Gallup Panel The Gallup Panel was created in 2004 as a proprietary, probability-based longitudinal panel of U.S. households that have been selected using random-digit-dial (RDD) sampling methods. Panel households are recruited through an outbound phone interview, and they agree to participate in an average of three surveys per month via phone, Web, or mail. Once in the panel, members are not required to spend a specific predetermined amount of time as panelists. Rather, they are encouraged to remain members as long as they are willing and interested. There are no incentives or financial rewards for participating in the panel, though several token thank you gifts are sent throughout the year. Lastly, as with any longitudinal design, the Gallup Panel is affected by attrition. To leave the Gallup Panel, members can call the tollfree support phone number and request removal, or they are removed from the panel after they fail to respond to six consecutive surveys (with a postcard prompt after the third miss). Monthly attrition rate averages between 2% and 3%. In addition to client-sponsored research, internal profile studies are conducted every three weeks with the entire adult panel population. These profile studies are designed to gather hundreds of behavioral, attitudinal, psychographic, and demographic statistics from the panelists. The data collected on these profile studies are used to target individuals for future custom research, to gain in-depth understanding of a particular industry or social issue, and to track longitudinal changes in panelist behavior and opinions. Copyright 2011 Gallup, Inc. All rights reserved. 21