I want to make a meta-analysis on household consumption expenditures elasticity data. In these studies that I've got; methods and regions and etc. varies according to studies. But at last, the results in these studies are seems to be comparable basis to elasticity. And these elasticities results are published in a single column data (could be thought according to COICOP statistical classification) in each article. So, from these studies, I have only data in columns according to expenditure patterns. Which meta-analysis method should i use to combine these studies, for example, can i use single-case meta-analysis methods (PND etc.). Or should I use systematic review method?
I put an example of data set in the attachment (in data set example, i wrote the same values for each study to be an example, but of course it's different in each study). How can i combine these studies (data from articles or thesis), is there a meta-analysis method to combine these studies data, or any other statistical methods? Because of the heterogeneity should i not use meta-analysis, or is the systematic review method the only suitable method for combining or quantitative synthesis of this data?
You need a measure of the uncertainty of elasticity (e.g. standard error, 95% confidence interval) to conduct a meta-analysis (and also to estimate heterogeneity as you mention) . A systematic review could include(not necessarily) a meta-analysis. They are not competitive methods.
Hi Gyanendra Kumar, income elasticity of demand measures the responsiveness of the demand for a good to a change in the income of the people demanding the good, ceteris paribus. It is calculated as the ratio of the percentage change in demand to the percentage change in income. For example, if, in response to a 10% increase in income, the demand for a good increased by 20%, the income elasticity of demand would be 20%/10% = 2. I think elasticity is a continuous variable at the same time.
Hİ Dimitris Mavridis, so can i get something from these column data of studies for meta-analysis (std. error etc.)? İf meta- analysis is not possible for this column datas, which method can i try to combine these sudies, If nothing can not be made for meta-analysis, does the systematic review the last option to apply?
You can not get estimates of uncertainty from the data you uploaded because you do not have within-study data. Perhaps you can get them from the original papers or the authors.
Finding those studies should have been a product of a systematic review (e.g. studies should satisfy pre-specified criteria, how and where you found these studies...). Meta analysis is a part (not necessarily though) of a systematic review.
I am sorry but I can not suggest any analyses as I am not familiar with that type of data and their analyses.
Hi Dimitris, as you noted, studies i have satisfies pre-specified criterias, namely i choiced these studies according to these criterias. But there arent any effect size statistics or any treatment and control groups design or any Pretest-Posttest-Control Group Designs in these studies, so only the systematic review seems the right choice here to combine the studies? What is your opinion?
If you want to meta-analyzed the elasticity, you probably can try the conventional meta-analysis method, i.e. the inverse-variance weighted method, provided that you can get estimate the SE from each study.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics