Discussion: View Thread

Dear colleagues,

I am working on a paper with a professor regarding testing a meta-analytic path model. There are six variables in the model. However, for some relationships, we do not have enough studies. We would greatly appreciate if you could provide us with published or unpublished studies that include the following relationships:

1. Anxiety and problem-focused coping

2. Worry and problem-focused coping

3. Challenge stressors and threat appraisal

4. Positive affect and problem-focused coping

Best Regards,

Pan

---

Liyao Pan, M.A.

Doctoral student

Department of Management

The University of Texas at San Antonio

Pan,

In many situations, like yours, it may not be possible to find studies to complete a targeted meta-analytically derived correlation matrix. As noted by Bergh et al. (in press, SMJ, listed below), possible solutions to this challenge include: "(1) search again for relevant studies based on the same and/or different search criteria; (2) contact other researchers to request correlation values; (3) conduct original primary-level studies; (4) replace some existing variables with conceptually similar 'surrogate' variables; (5) group specific constructs into broader construct(s) by deriving composite correlations; (6) use a two-stage structural equation modeling, which manages missing values automatically; (7) implement advanced data imputation techniques such as full information maximum likelihood; (8) use the average effect size across all non-missing effect sizes; or (9) rely on subject matter experts or expertise to estimate the value for the missing effect sizes." Bergh et al recommended the first three options as the most desirable because all of them involve using actual data without any conceptual or empirical manipulation. If none of the above possible solutions are available, the remaining option is to limit the scope of inquiry. For more detailed information on these issues, see:

·       Bergh, D. D., Aguinis, H., Heavey, C. Ketchen, D. J., Boyd, B. K., Su, P., Lau, C., & Joo, H. in press. Using meta-analytic structural equation modeling to advance strategic management research: Guidelines and an empirical illustration via the strategic leadership-performance relationship. Strategic Management Journal. doi: 10.1002/smj.2338 [available at http://mypage.iu.edu/~haguinis/pubs.html]

I hope this helps!

All the best,

--Herman.