The previous article explained how to enter data for performing a meta-analysis of the studies reporting comparison data. Therefore this article explains the data entry methodology, for performing a meta-analysis on the outcomes of studies having a single group. For this purpose, the single group of patients who were administered drug treatment will be taken into consideration.
One group and matched group designs in Meta-analysis
In matched group design, the unit of analysis is the pair of participants who have been matched on the basis of a particular attribute as deemed fit by the researcher. One member from each pair is assigned to different groups, thus each pair serves as its own control. This tends to reduce error and increase statistical significance of results. Here, the degree of correlation has particular importance, because it influences the degree of impact between the matched participants (White & Sabarwal 2014).
One group (pre-post) design includes administration of the treatment to a single group of patients. Also, the evaluations are carried out by measuring the patient condition before and after the treatment (Marsden & Torgerson 20152). In the present Obsessive-Compulsive Disorder (OCD) case study, the patients were administered drug treatment and their pre-post treatment scores were measured over the CY-BOCS scale.
One group (pre-post) data entry
The figure below shows the effect size data directory for ‘One group (Pre-post) and matched groups’. The encircled format is the desired format for the present demonstration. Select any of the formats as per their preference and available data.
The figure below shows the interface of the spreadsheet pertaining to the ‘Mean difference, Standard Deviation of difference and sample size’ format in the ‘One group (pre-post)’ directory. However, this format offers no option for customizing the field names and the columns have already been named as per the data to be entered. Also, the effect measures to be calculated are the same as in the previous article as the outcome is continuous.
- Mean difference: It is obtained by carrying out simple subtraction of the post mean from the pre mean.
- The standard deviation of difference: Here the standard deviation of the difference scores has to be calculated. It could be obtained by using the following formula:
Here also ‘Pre or Post Corr’ field refers to the correlation values between the pre and post-scores within the group. The correlation values were not readily available, hence imputed (0.5), following the same logic as in the previous article.
The figure below shows the completed spreadsheet, with the calculated effect size. The mean difference values have been kept as negative, as the post scores were lower than the pre-scores. The effect direction was determined automatically by the sign of ‘Mean difference’.
Data analysis procedures
Furthermore, proceed with the analysis for the single group studies. The next article will discuss data analysis procedures, wherein the results for both the case studies will be analyzed and discussed.