meta analysis

Question 1

what is a meta analysis?

Answer 1

• the quantitative synthesis (merging) of studies that are similar to eachother
• usually conducted when there is a huge body of research or where the research is inconsistent
• can include research from different cultures, times and places
• focuses on the patterns and relationships between factors/concepts being investigated

Question 2

what are strengths of using a meta analysis?

Answer 2

• more generalisable, as they use large samples and include) data from different cultures, places and times
• allows reliability, as they can uncover contradictory findings and where research tends to agree
• ethical - researchers examine data from previous studies, using secondary data
• objective - only uses quantitative data, and forms statistical analysis of combined data
• (robust data) best evidence synthesis, where data has already been double checked and analysed = high valiity

Question 3

what are weaknesses of using a meta analysis?

Answer 3

• small studies use a lot of different designs which affects the data we collect = becomes part of a confounding variabe
• experimenter bias - a person may choose to analyse studies that are most relevant to their own hypothesis
• publication bias - where we only use studies that had success rates
• there are risks of using unpublished data that hasn’t been peer reviewed (unchecked methodology or calculation) so any errors in that research is will occur in our meta analysis
• objective data analysis = lacks context and becomes a flaw when analysing trends and patterns

Question 4

what is the process of a meta analysis?

Answer 4

1. a research question is identified and a hypothesis is proposed
2. a systematic review is specifically designed to address the research question to identify all studies considered to be both relevant and of sufficiently good quality.
3. once data is selected, summarised data or outcomes are taken from each study - depending on the study and the research question, outcome measures could be numerical or categorical (the analysis is statistical)
4. studies then need to be standardised to produce
   estimates to make effect sizees.
5. draw conclusions and state trends founf and the effect size of such strends