meta analysis Flashcards
(4 cards)
what is meta analysis
a way of using results from different studies, about the same issue, and studying them as a whole to look for an overall picture about the area of study- combining stats tests/ analysing.
If a number of studies separately find the same answer, and then those studies are analysed together (a meta-analysis) then that answer becomes stronger as the studies support one another.
Primary data -Information collected directly from first hand experience. Data collected specifically for a specific study. e.g. questionnaires, an experiment.
Secondary data -Information that was collected for a purpose other than the current one. e.g. government statistics, results from a different experiment.
stages / how to conduct a meta analysis
Define the research question: consider the size and select a manageable topic but make sure it’s large enough to conduct a meta analysis, often involves examining types of effects for example associations between variables
Perform literature search
- identify, retrieve and review published and unpublished studies be exhaustive online databases such as psych info, pay review journals books and dissertation
Select the studies,
Set up a select and reject scheme, read the titles and abstract of relevant research papers
Once identified a certain number of studies that must be included, obtain the full text and select and reject again
Extract data
- create table spreadsheets, et cetera
Analyse the data
-using statistical software calculate an effect size
Allowing comparison of statistical differences between groups
Report results
- summarising findings clearly and transparently presented as tables graphs and a narrative format highlighting key trends and the implications
strengths
Use a secondary data, therefore the researcher is not breaching any ethical guidelines as it can be presumed that ethical guidelines were followed initially
High validity as conclusions are drawn on a wide sample of participants including a range of cultures, ages occupations and reducing anomalies which may be average out allowing a cause in effect relationship
Multiple studies gathering high death and detail and credibility
weaknesses
Using secondary data may lack validity and reliability as maybe unaware of potential bosses that originally occurred for example issues with control and dies procedure operationtionisation
Maybe less valid because it’s been gathered previously for a different purpose which could be quite different to the purpose of the current research
Secondary data may have been collected some time ago therefore may not be temporarily valid
Publication bias positive studies are more likely to go to print and search for the search for studies can produce unintentionally bias results for example using an incomplete set of words or varying strategies to search databases
