Meta-analysis

Question 1

AO1/Description

Answer 1

• We often use these in clinical to get a whole picture about the effectiveness of treatment. A meta-analysis is a statistical analysis that combines the results of multiple scientific studies.
• You collect a sample of research into a set topic and then analyse all of the data with a new set of statistical tests to get a larger overview of statistical significance.
• Meta-analysis means an analysis over others (an analysis of analyses basically) and is a way of using results from different studies, about the same issue, and studying them as a while to look for an overall picture about the area of study.
• IMPORTANTLY this usually involves doing more statistics
• If a number of studies separately find the same answer, and then those studies are analysed together then the answer becomes stronger as the studies back one another up (and we get a whole overview.

Question 2

AO3: Strengths

Answer 2

• It would be much quicker to use secondary data as she would not need to design and carry out her own research as data already exists
• The use of secondary data is also less expensive as the costs involved in producing a questionnaire do not have to be incurred, the analysis can be done directly
• Does allow for a further analysis which lets us get an overview of patterns which single studies cannot (gives us better data to make conclusions about)
• Meta-analysis allow us to gather large samples
• Statistical analysis can be conducted to see if results are significant
• Large sample sizes can be gathered which would be impossible otherwise

Question 3

AO3: Weaknesses

Answer 3

• Publication bias is a factor (non-significant results are not published…which skews and biases the data)
• Secondary data may be less satisfactory as cannot be sure how the previous student collected the data so it may not be appropriate for her study
• Secondary data may have been collected some time ago so responses may have changed as views in society are constantly changing (time locked/lacks temporal validity)
• Bias in what is selected for the meta-analysis