0.5 Mills, l., mcdowelle, j., & Roush, w. (july 2009). does research  (Page 4/7)

A meta-analysis is defined as “the statistical analysis of a large collection of analysis results from individual results for the purpose of integrating the findings” (Glass, McGaw&Smith, 1981, p.3). Hamer and Simpson (2002) state that meta-analysis is a particularly useful tool for synthesizing and integrating large quantities of information from many studies with contradictory information. Since a large body of contradictory information exists on the question of whether emotional intelligence does have an effect on effective leadership it was thought a meta-analysis examining the relationship between emotional intelligence and effective leadership would be helpful in answering many of the questions surrounding this newly popular concept. A meta-analysis of this relationship can help those responsible for school leadership determine if the skills associated with Emotional Intelligence should be an important element in the evaluation of school leaders.

Results of the study

The comprehensive attempt to identify relevant studies on EI’s effect on leadership effectiveness conducted between 1990 to the present yielded 141 studies. Forty-eight studies met the criteria for inclusion in the meta-analysis sample. Those 48 studies yielded 99 correlations for analysis. Over half of the studies included in this study were in the form of unpublished dissertations (48%) and theses (8%) with forty-one percent of participants reported serving in the field of business or industry.

Determining the degree of homogeneity of the studies included in a meta-analysis helps to determine the statistical model to be used. A Q statistic was computed to explore whether the variability across effect sizes was greater than expected from sampling error alone. Based on a significant result for the Q statistic for these data, the hypothesis of homogeneity was rejected and a random-effects model was implemented. This significant Q statistic result was not unexpected given the essential random differences between studies that were associated with the framework of EI chosen and other study variables (e.g. the framework of leadership effectiveness, occupational setting, etc.). Incorporation of a random effects model allows for the estimation of the mean of a distribution of effects which prevents the underestimation of the weight of a small study or the overestimation of the weight of a large study (Borenstein, Hedges, Higgins,&Rothstein, 2007).

Using Biostat’s Comprehensive Meta-Analysis Software Version 2.0, the 99 effect sizes from the correlational studies with a total of 7,343 subjects were converted into Fisher z scores and an overall effect size z was computed and converted to the initial metric of correlation coefficient. Based on the rule of thumb for product moment correlation effect size magnitudes suggested by Lipsey and Wilson (2001), correlation effect size values are considered small if less than or equal to .10, medium if equal to .25, and large if greater than or equal to .40. This meta-analysis yielded a combined effect of r= .383 which can be interpreted as a moderately strong relationship between emotional intelligence and leadership effectiveness.