Question

Which statement about the design of a meta-analytic study is true?

	A.	The quality of a report is not taken into consideration in a meta-analysis.
	B.	If statistical heterogeneity is anticipated, subgroup analyses to explore systematic differences in effects should be planned.
	C.	An important exclusion criterion in a meta-analysis is the low quality of a study.
	D.	Unpublished research reports are excluded from meta-analyses.

Answer 1

The quality of a report is not taken into consideration in a meta-analysis - False

The studies used for meta analyses should have maintained a standard quality in following the methodology, to avoid un due bias in the results.

If statistical heterogeneity is anticipated, subgroup analyses to explore systematic differences in effects should be planned –True

Statistical heterogeneity is a situation where there is presence of observed variability between study estimates beyond the expectation. Sub group analyses is an exploratory analysis, done either by grouping the subsets of population or by grouping subsets of studies whereby different studies would be in each subgroup.

An important exclusion criterion in a meta-analysis is the low quality of a study – False

Exclusion of the studies considered to be of low quality, may results in missing certain information and data from the excluded study, which may not be present in the other studies. The study is been termed of low quality, as it has travelled in a different direction than the other studies. Therefore including only the studies with high quality score may provide us result, which is a mere projection of those studies.

Unpublished research reports are excluded from meta-analyses –False

Unpublished studies in general cannot be considered as a criteria for exclusion, also a work unpublished does not mean it is invalid. Any study with adequate empiric facts could be included in any research work. But one cannot deny a fact that published research is more likely to contain results that are statistically significant. Though the estimated effect size of published research are larger than of unpublished studies, the methodological rigor doesn’t differ among the both.