Mega-Analysis of Culture: Synergy of Network
 |
 |
 |
 |
 |
 |
MEGA-ANALYSIS OF EFFECT OF CULTURE
The mega-analytic project stems from a series of meta-analyses and has two main purposes:
1. Re-evaluation of national cultural indices (similar to those offered by Hofstede) and analysis of
cultural change;
2. Evaluation of effects of culture and culture-related self-construals in organizational and societal
contexts.
The meta-analytic dataset contains data from total of 781 empirical papers that were based on Hofstede’s
framework of culture or subsequent related models of cultural values and self-construals. The dataset
summarizes responses of over a million individuals from 64 countries.
While meta-analysis is a powerful and increasingly popular data analysis technique, it has numerous
limitations as it is based on sample-level data. Absence of individual-level data makes it very difficult to
establish empirical equivalence of measures and psychometric properties of data included in a meta-analytic
dataset. As a result, meta-analyses are often criticized for limited commensurability across studies in meta-
analytic samples, aka the “apples and oranges” problem. Furthermore, inability to analyze individual
responses seriously limits the types of research questions that can be addresses with a meta-analysis.
When coding studies for the meta-analysis, we had to contact dozens of scholars with requests for
additional descriptives that were not reported in some papers. To our surprise, many sent us their complete
original individual-level datasets, which lead to the idea of the mega-analysis.
Mega-analysis is not an established research technique and even the term “mega-analysis” is completely
new. Basically, the plan is to combine all original individual-level datasets from the 781 studies, bring the
data to common metric and using these refined individual-level data revisit the research questions originally
addressed in the individual studies included in our dataset. The mega-analytic dataset would also provide a
either separate individual-level datasets, or meta-analytic sample-level data, such as those related to
moderating effects, longitudinal perspective, and instrument property re-evaluation.
We are now inviting the authors of the studies in our dataset to join us as co-authors on the project and pool
their individual datasets. The meta-analytic dataset represents responses from over a million individuals, so
even a fraction of individual datasets would yield a combined sample size of thousands of observations. We
have already contacted number of scholars and the response has been exceptionally enthusiastic.
RESEARCH QUESTIONS
Four types of research questions will be addressed using the mega-analytic individual-level dataset:
1. Re-evaluation of effects of culture on various organizational processes and outcomes;
2. Moderator analysis;
3. Re-evaluation of national cultural means and analysis of cultural change;
4. Comparative analysis and re-evaluation of psychometric properties of the included instruments.
Re-evaluation of effects of culture on various organizational processes and outcomes
The studies in our database explore effects of culture in 60 categories (Appendix 1). The relationships in
each category have been studied in 10 to 200 studies. A mega-analytic combined dataset would allow for
re-evaluating the relationships more accurately with pooled original individual-level data from each of the
studies. We are planning to write a general mega-analytic review paper on effects of culture in
organizational context and a series of individual papers focusing on specific categories of relationships.
Moderator analysis
In addition to main variables (culture and organizational processes and outcomes), the pooled mega-analytic
dataset also contains data on individual sample characteristics, research design of each study, as well as a
range of secondary variables. This would allow for testing various moderation and mediation effects and
assessing generalizability of findings across different contexts and settings.
Re-evaluation of national cultural means and analysis of cultural change
In addition to correlational research, mega-analysis allows for re-evaluation of national cultural means, much
like those reported by Hofstede and other scholars. A much larger and more representative mega-analytic
dataset would allow for more precise estimates of national and regional scores and rankings. Moreover,
since different studies included in the dataset were conducted at different times, the pooled dataset would
allow for longitudinally explore cultural change.
Comparative analysis and re-evaluation of psychometric properties of the instruments
Finally, the mega-analysis allows for re-evaluating psychometric properties, factor structures, and cross-
cultural generalizability of various construct measures, as well as analysis of the relationship between
various alternative measures and constructs, overlap and interchangeability among them, and their criterion
validity and relationship with other external contracts.
PARTICIPATION IN THE MEGA-ANALYTIC PROJECT
The authors of the individual studies in our database are invited to join us at the mega-analytic project as co-
authors. The participation entails sharing the original individual-level datasets (sample descriptives and
individual responses to each instrument item), as well as optional collaboration at developing and writing
resulting paper(s). Every data contributor will be added as a co-author of the resulting publications that
were directly based on the data contributed by the scholar, unless otherwise agreed. The order of
authorship will be determined by the individual input in each resulting publication. More detailed instructions
will be provided upon your response.
APPENDIX 1.
OUTCOME CATEGORIES EXPECTED TO BE REPRESENTED IN THE MEGA-ANALYTIC DATASET
1. Absenteeism
2. Adjustment (to new environment)
3. Affect (negative, positive)
4. Anxiety
5. Attribution bias (internal/external)
6. Bias (evaluation, in-group favoritism)
7. Cohesiveness
8. Commitment (organizational, team)
9. Communication (indirectness; self-promotion; sensitivity to others; politeness; interruption rate;
ritualism, etc)
10. Competitiveness
11. Concern for interest of others vs. concern for own interest
12. Conflict management (avoidance, competition, accommodation, compromising; preference for third
party involvement)
13. Conformity
14. Cooperation (cooperation in teams; during negotiations
15. Coping (active, passive)
16. Corruption (perceptions, propensity to engage in)
17. Counseling (willingness to seek counseling, etc)
18. Depression
19. Economy (economic growth, wealth, etc)
20. Self-efficacy
21. Effort
22. Shame, embarrassibility
23. Emotions (tendency to display emotions, positive/negative)
24. Entrepreneurship
25. Ethics sensitivity, avoiding unethical behavior
26. Equality (equalitarian values/behaviors)
27. Exchange ideology (this-for-that)
28. Family values, importance of family
29. Feedback (seeking, perceived accuracy of)
30. Gender role equality
31. Self-perception (identification with organization, region/country)
32. Innovativeness
33. Cognitive ability
34. Justice (reaction to injustice; perceived justice in organization)
35. Leadership style preference (charisma; directive; participative; inspiring/supportive;
transformational; preference for leader who stimulates intellectual input)
36. Preferences for different learning styles
37. Locus of control
38. Management (preference for empowerment, self-management, paternalism, informal HRM system)
39. Marketing (commitment to product, company, service provider)
40. Rewards (preference for different types of rewards)
41. Org design (centralization, job complexity, formalization, unit differentiation, organic/mechanic)
42. Mental health
43. Motivation
44. Negotiation behavior (compromising; negotiation outcome, joint/individual profit; propensity to plan,
defensiveness; satisfaction with opponent behavior, etc)
45. Org. citizenship behavior
46. Performance (various individual and group level measures)
47. Personality (Agreeableness, Conscientiousness, Extraversion, Neuroticism, Openness)
48. Sensitivity to rejection
49. Recruitment (preference for different modes)
50. Selection (rigor and extent of verification)
51. Concern for relationship and relationship harmony
52. Religiosity
53. Risk liking
54. Attitudes to rules (particularism, universalism)
55. Satisfaction (supervisor, job, life, coworker)
56. Self-esteem
57. Sexual orientation: homosexuality
58. Social factors (need for affiliation, social avoidance, social independence)
59. Stress
60. Teamwork preference
61. Trust
The mega-analytic project stems from a series of meta-analyses and has two main purposes:
1. Re-evaluation of national cultural indices (similar to those offered by Hofstede) and analysis of
cultural change;
2. Evaluation of effects of culture and culture-related self-construals in organizational and societal
contexts.
The meta-analytic dataset contains data from total of 781 empirical papers that were based on Hofstede’s
framework of culture or subsequent related models of cultural values and self-construals. The dataset
summarizes responses of over a million individuals from 64 countries.
While meta-analysis is a powerful and increasingly popular data analysis technique, it has numerous
limitations as it is based on sample-level data. Absence of individual-level data makes it very difficult to
establish empirical equivalence of measures and psychometric properties of data included in a meta-analytic
dataset. As a result, meta-analyses are often criticized for limited commensurability across studies in meta-
analytic samples, aka the “apples and oranges” problem. Furthermore, inability to analyze individual
responses seriously limits the types of research questions that can be addresses with a meta-analysis.
When coding studies for the meta-analysis, we had to contact dozens of scholars with requests for
additional descriptives that were not reported in some papers. To our surprise, many sent us their complete
original individual-level datasets, which lead to the idea of the mega-analysis.
Mega-analysis is not an established research technique and even the term “mega-analysis” is completely
new. Basically, the plan is to combine all original individual-level datasets from the 781 studies, bring the
data to common metric and using these refined individual-level data revisit the research questions originally
addressed in the individual studies included in our dataset. The mega-analytic dataset would also provide a
either separate individual-level datasets, or meta-analytic sample-level data, such as those related to
moderating effects, longitudinal perspective, and instrument property re-evaluation.
We are now inviting the authors of the studies in our dataset to join us as co-authors on the project and pool
their individual datasets. The meta-analytic dataset represents responses from over a million individuals, so
even a fraction of individual datasets would yield a combined sample size of thousands of observations. We
have already contacted number of scholars and the response has been exceptionally enthusiastic.
RESEARCH QUESTIONS
Four types of research questions will be addressed using the mega-analytic individual-level dataset:
1. Re-evaluation of effects of culture on various organizational processes and outcomes;
2. Moderator analysis;
3. Re-evaluation of national cultural means and analysis of cultural change;
4. Comparative analysis and re-evaluation of psychometric properties of the included instruments.
Re-evaluation of effects of culture on various organizational processes and outcomes
The studies in our database explore effects of culture in 60 categories (Appendix 1). The relationships in
each category have been studied in 10 to 200 studies. A mega-analytic combined dataset would allow for
re-evaluating the relationships more accurately with pooled original individual-level data from each of the
studies. We are planning to write a general mega-analytic review paper on effects of culture in
organizational context and a series of individual papers focusing on specific categories of relationships.
Moderator analysis
In addition to main variables (culture and organizational processes and outcomes), the pooled mega-analytic
dataset also contains data on individual sample characteristics, research design of each study, as well as a
range of secondary variables. This would allow for testing various moderation and mediation effects and
assessing generalizability of findings across different contexts and settings.
Re-evaluation of national cultural means and analysis of cultural change
In addition to correlational research, mega-analysis allows for re-evaluation of national cultural means, much
like those reported by Hofstede and other scholars. A much larger and more representative mega-analytic
dataset would allow for more precise estimates of national and regional scores and rankings. Moreover,
since different studies included in the dataset were conducted at different times, the pooled dataset would
allow for longitudinally explore cultural change.
Comparative analysis and re-evaluation of psychometric properties of the instruments
Finally, the mega-analysis allows for re-evaluating psychometric properties, factor structures, and cross-
cultural generalizability of various construct measures, as well as analysis of the relationship between
various alternative measures and constructs, overlap and interchangeability among them, and their criterion
validity and relationship with other external contracts.
PARTICIPATION IN THE MEGA-ANALYTIC PROJECT
The authors of the individual studies in our database are invited to join us at the mega-analytic project as co-
authors. The participation entails sharing the original individual-level datasets (sample descriptives and
individual responses to each instrument item), as well as optional collaboration at developing and writing
resulting paper(s). Every data contributor will be added as a co-author of the resulting publications that
were directly based on the data contributed by the scholar, unless otherwise agreed. The order of
authorship will be determined by the individual input in each resulting publication. More detailed instructions
will be provided upon your response.
APPENDIX 1.
OUTCOME CATEGORIES EXPECTED TO BE REPRESENTED IN THE MEGA-ANALYTIC DATASET
1. Absenteeism
2. Adjustment (to new environment)
3. Affect (negative, positive)
4. Anxiety
5. Attribution bias (internal/external)
6. Bias (evaluation, in-group favoritism)
7. Cohesiveness
8. Commitment (organizational, team)
9. Communication (indirectness; self-promotion; sensitivity to others; politeness; interruption rate;
ritualism, etc)
10. Competitiveness
11. Concern for interest of others vs. concern for own interest
12. Conflict management (avoidance, competition, accommodation, compromising; preference for third
party involvement)
13. Conformity
14. Cooperation (cooperation in teams; during negotiations
15. Coping (active, passive)
16. Corruption (perceptions, propensity to engage in)
17. Counseling (willingness to seek counseling, etc)
18. Depression
19. Economy (economic growth, wealth, etc)
20. Self-efficacy
21. Effort
22. Shame, embarrassibility
23. Emotions (tendency to display emotions, positive/negative)
24. Entrepreneurship
25. Ethics sensitivity, avoiding unethical behavior
26. Equality (equalitarian values/behaviors)
27. Exchange ideology (this-for-that)
28. Family values, importance of family
29. Feedback (seeking, perceived accuracy of)
30. Gender role equality
31. Self-perception (identification with organization, region/country)
32. Innovativeness
33. Cognitive ability
34. Justice (reaction to injustice; perceived justice in organization)
35. Leadership style preference (charisma; directive; participative; inspiring/supportive;
transformational; preference for leader who stimulates intellectual input)
36. Preferences for different learning styles
37. Locus of control
38. Management (preference for empowerment, self-management, paternalism, informal HRM system)
39. Marketing (commitment to product, company, service provider)
40. Rewards (preference for different types of rewards)
41. Org design (centralization, job complexity, formalization, unit differentiation, organic/mechanic)
42. Mental health
43. Motivation
44. Negotiation behavior (compromising; negotiation outcome, joint/individual profit; propensity to plan,
defensiveness; satisfaction with opponent behavior, etc)
45. Org. citizenship behavior
46. Performance (various individual and group level measures)
47. Personality (Agreeableness, Conscientiousness, Extraversion, Neuroticism, Openness)
48. Sensitivity to rejection
49. Recruitment (preference for different modes)
50. Selection (rigor and extent of verification)
51. Concern for relationship and relationship harmony
52. Religiosity
53. Risk liking
54. Attitudes to rules (particularism, universalism)
55. Satisfaction (supervisor, job, life, coworker)
56. Self-esteem
57. Sexual orientation: homosexuality
58. Social factors (need for affiliation, social avoidance, social independence)
59. Stress
60. Teamwork preference
61. Trust
