Data Management in Meta-analytical Procedures: Myths, Realities, and Biases

• Author: Popa, R. I.
• Position: Faculty of Psychology and Education Sciences, University of Bucharest
• Pages: 175-180

Bulletin of the Transilvania University of Braşov • Vol. 5 (54) No. 1 - 2012

Series VII: Social Sciences • Law

DATA MANAGEMENT IN META-

ANALYTICAL PROCEDURES:

MYTHS, REALITIES AND BIASES

Radu I. POPA1

Abstract: The aim of the present paper is to highlight the importance of a

correct data management and the preparation of a proper statistical value

base in developing a meta-analysis. When dealing with meta-analytical

procedures, research studies encounter a wide range of challenges, ranging

from the classical value losses towards more modern coding or sampling

biases. The theoretical and practical examples of this paper indicate that a

meta-analytical approach is most useful and applicable when related to a

clear problem specification, a correct coding scheme and the use of eligible

study reports.

Key words: meta-analysis, coding protocol, study descriptors, bias.

1 Faculty of Psychology and Education Sciences, University of Bucharest.

1. Introduction

Across the long history of methods

inside the psychological field of research,

many a time we have witnessed the lack of

a quantitative procedure which could

revise a research theme or phenomenon, in

order to determine in a certain degree the

result of a study, which can be replicated

by other studies with success.

In 1952, Hans Eysenck [6] conducted a

challenging debate upon the beneficial

effects of psychotherapy on patients, while

Gene V. Glass offered a statistically

standardized response called ‘meta-

analysis’ in order to present the

effectiveness of psychotherapy a cross a

wide range of studies [6]. Since 1970

meta-analysis has been de veloping as a

collection of methods and techniques for

quantitative research synthesis.

In time, the meta-analysis approach has

been refined, its methodology being

restructured and applied to various

subjects. This form of survey or integrated

analysis still offers answers to research

reports on psychological processes at a

complex level, with a strong possibility to

generalize results, theories and paradigms.

2. A theoretical view on meta-analysis

Meta-analysis procedures impose a

structured approach on the evaluation of

research findings. It brings out a much

more complex and strong apparatus,

resolving a wide range of issues, which are

dealt in the conventional r esearch review

techniques [5]. Specification of criteria,

coding strategies for eligible studies, data

analysis and protocols, sampling

information and possible biases are a few

elements which recommend the meta-

analysis as a proper method when