Formulate problem
Relevant theoretical and empirical articles
Riketta (2008)
The focal problem of the research is the interconnection between job attitudes and performance (Riketta, 2008). The author refers to the relevant theoretical articles: Brief and Weiss (2002), Cooper-Hakim and Viswesvaran (2005), etc.
Harrison, Newman, and Roth (2006)
The researchers refer to a series of theoretical articles: Austin and Villanova, Judge, Thoreson, Bono, and Patton, etc (Harrison, Newman, & Roth, 2006). Many sources are rather dated.
Important, common variables, study, and measurement characteristics to assess
Riketta (2008)
The common variables are improved work outcomes. In other words, the examined studies mainly assess the relation between positive job attitudes and improved performance. The positive attitudes are subdivided into such variables as job satisfaction, job commitment. The performance is subdivided into such variables as in-role and extra-role performance.
Harrison, Newman, and Roth (2006)
The common variables are general attitudes, job satisfaction, and supervisor ratings.
Contradictory findings
Riketta (2008)
There are varied interpretations of the attitude-performance relations. The first model implies that the quality of performance is determined by the job attitudes. The second model implies that job attitudes are shaped by performance. The third model implies that these two variables are interconnected and equally impact one another. According to the fourth model, the two variables are not interdependent.
Harrison, Newman, and Roth (2006)
Some researchers believe that job attitudes determine mainly such factors as absence and turnover, while others assume that they impact in-role performance as well.
Meta-analysis objectives
Riketta (2008)
The analysis is aimed at investigating the type of job attitude-performance relation. The exclusiveness of this meta-analysis resides in the fact that it examines the described type of relations in the framework of the time lag context. In other words, it tries to assess the potential impact of various time lags.
Harrison, Newman, and Roth (2006)
The researchers aim to test the hypothesis that job satisfaction and commitment are the key predictors of performance.
Preliminary theoretical framework
Riketta (2008)
The analysis relies upon the theory, according to which, that the applied timeframe determines how close the potential interconnection between job attitudes and performance might be. It is assumed that a longer timeframe implies stronger bonds between the examined variables.
Harrison, Newman, and Roth (2006)
The analysis relies upon the theory proposed by Judge which implies that job satisfaction and commitment reflect the general job attitude and shape the performance.
Collect Data
Conduct key-word database searches
Riketta (2008)
The following key-words were used in the process of database searching: satisfaction, commitment, work, job, organization, cross-lagged, longitudinal, in-role, extra-role performance, citizenship, work motivation, effort, and productivity.
Harrison, Newman, and Roth (2006)
The researchers used the following keywords: contextual performance, absenteeism, satisfaction, commitment, etc.
Search of references for key studies
Riketta (2008)
The author reports that all the references of the key studies were likewise examined.
Harrison, Newman, and Roth (2006)
The researchers report that they manually searched for references to key studies for absenteeism.
Manual search of relevant publications
Riketta (2008)
The author does not report whether any manual search was carried out.
Harrison, Newman, and Roth (2006)
The researchers performed a computerized search for the relevant publications.
Solicit studies from key researchers
Riketta (2008)
The key researchers were contacted even if their studies did not meet the inclusion criteria. In this case, they were asked to provide the missing data.
Harrison, Newman, and Roth (2006)
The researchers did not mention any solicit studies from key researchers.
Requests for articles and working papers on academic list-serves
Riketta (2008)
Several requests were posted to the Academy of Management and the Society for Industrial and Organizational Psychology.
Harrison, Newman, and Roth (2006)
The researchers did not post any additional requests.
Elimination criteria
Riketta (2008)
The author suggests five elimination criteria. The employees are excluded if they do not work for an organization. All other types of job attitudes but satisfaction and commitment are excluded. Those organizations that experienced a recent change are excluded. Group-level data is excluded. Studies that fail to report all the target correlations are excluded.
Harrison, Newman, and Roth (2006)
The researchers do not have a consistent elimination criteria model, though they designed several separate models with the relevant criterion for each.
Screen studies according to elimination criteria
Riketta (2008)
The studies that did not meet at least one of the set criteria were excluded from the described meta-analysis.
Harrison, Newman, and Roth (2006)
The researchers did not set any strict frame to screen the studies, they rather overviewed the general scope of the data collected.
Sufficiency of data to proceed with meta-analysis
Riketta (2008)
The researcher reports that the scope of the data collected in the course of the literature review was excessive.
Harrison, Newman, and Roth (2006)
The researchers agreed that they collected a sufficient amount of data when they had the data that met the criterion set for each of the designed models.
Evaluate Data
Common metrics across studies (e.g. correlation coefficient)
Riketta (2008)
The average size of the studies was relatively similar – hence, the sample would normally comprise 192 respondents. Most of the studies were carried out in English-speaking countries.
Harrison, Newman, and Roth (2006)
The researchers fail to identify any common metrics.
Key substantial and methodological study characteristics
Riketta (2008)
The major part of the studies employed the Job descriptive Index to measure the job satisfaction variable. Moreover, most studies used the Organizational Commitment Questionnaire to assess the job commitment variable. The in-role and extra-role variables were mainly measured with the help of self-reports.
Harrison, Newman, and Roth (2006)
The major part of the studies employs large samples and has a non-experimental design.
The coding form and manual
Riketta (2008)
The researcher designed a special coding table to organize the analyzed studies.
Harrison, Newman, and Roth (2006)
The researchers use a coding form designed by their colleagues.
Coding of each study
Riketta (2008)
The researcher coded 16 studies. On the whole, the coding table shows 6 types of coded relations.
Harrison, Newman, and Roth (2006)
The described form was used to code 514 studies.
Preparation of database and check for errors
Riketta (2008)
The researcher reports that all the information included in the database was thoroughly checked upon potential errors and its reliability. The described meta-analysis employed the random-effects model to determine the relevant standard deviation rate.
Harrison, Newman, and Roth (2006)
The researchers checked their databases as they provide relevant sampling error rates.
Analyze Data
Univariate analysis of mean effect sizes
Riketta (2008)
The researcher analyzed mean effected sizes, according to which, the average effect size for the entire study might be presented as ß=.00.
Harrison, Newman, and Roth (2006)
The researchers report the analysis of mean effect sizes in the form of a table.
Search for moderators using bivariate analysis of effect sizes
Riketta (2008)
The researcher pointed out such moderatos as job attitude, performance type that were searched for with the help of the bivariate analysis of effect sizes.
Harrison, Newman, and Roth (2006)
The researchers do not provide the findings of the bivariate analysis of effect sizes.
Search for moderators using multivariate analysis of effect sizes
Riketta (2008)
While applying the multivariate analysis of effect sizes, the researcher searched for such moderators as job attitude, performance, and time lag.
Harrison, Newman, and Roth (2006)
The researchers search for such moderators as contextual performance, compatible job attitude, etc. in the course of the multivariate analysis of effect sizes.
Analysis of the meta-analytical correlation matrix to test theory
Riketta (2008)
The researcher constructed the meta-analysis correlation matrix and used it to carry out the relevant regression analysis.
Harrison, Newman, and Roth (2006)
The researchers report that their equitation models rely upon the relevant correlation matrix.
Interpret and discuss
Report of univariate and bivariate findings using tables and figures
Riketta (2008)
The researcher reports the findings in the form of tables. No figures are used to illustrate the results.
Harrison, Newman, and Roth (2006)
The researchers use both tables and figures to report the findings.
Recommendations for future research based on the findings
Riketta (2008)
The researcher offers several general recommendations (ex. it is advised to use the panel design enlarge the database), though no detailed guidelines are provided.
Harrison, Newman, and Roth (2006)
The researchers provide detailed recommendations for future research: they suggest using conditional probability and stage sequential analysis instead of the correlational analysis and provide the rationale for this recommendation.
Critical Analysis
The information represented above shows the comparative analysis of two meta-analysis studies carried out by Riketta and Harrison. The analysis was performed relying upon the model designed by Kirca and Yaprak. This model implies that a consistent meta-analysis should have five stages (Kirca, & Yaprak, 2010).
Hence, Riketta follows all the guidelines suggested by Kirca and Yaprak. Meanwhile, the researcher fails to generate detailed recommendations for those who will decide to continue the research. So it might be recommended that the provided guidelines are supported with the relevant rationale to make the conclusion part more convincing (Swailes, 2002).
On the contrary, Harrison and the co-authors do not fulfill some requirements implied by Kirca’s model in the course of data collection. Thus, it is unclear which criteria they applied to include the studies in the meta-analysis. It might be recommended that they provided a more detailed description of the data collection stage. In the meantime, the researchers offer a comprehensive conclusion and give useful recommendations for their colleagues that might want to broaden this research.
Reference List
Harrison, D. A., Newman, D. A., & Roth, P. L. (2006). How important are job attitudes? Meta-analytic comparisons of integrative behavioral outcomes and time sequences. Academy of Management Journal, 49(2), 305-325.
Kirca, A. H., & Yaprak, A. (2010). The use of meta-analysis in international business research: Its current status and suggestions for better practice. International Business Review, 19(1), 306-314.
Riketta, M. (2008). The causal relation between job attitudes and performance: a meta-analysis of panel studies. Journal of Applied Psychology, 93(2), 472-481.
Swailes, S.F. (2002). Organizational commitment: a critique of the construct and measures. International Journal of Medical Reviews, 4(2), 155-178.