Hard Reserach Methods Flashcards
What is content analysis?
An indirect way of studying behaviour by looking at (observing) what humans create e.g. analysing posters, films or magazines. It involves generating themes or codes to analyse the behaviours behind the material - normally qualitative data.
- identify important categories from the study and then work through the source and tally the number of occurrences of each categories (draw a tally chart for an example)
What is investigator bias and how can you prevent it?
This is any influence the investigator/experimental situation has on ptps that might alter the results, that is not intended.
- ptps may not act naturally if they know they are being watched
- the investigators mood or demeanour may encourage or discourage ptps
—> can be avoided by using a single blind study or by delivering instructions in a standardised way (possible to do it via computer)
How could the reliability of content analysis be assessed?
Inter - rater reliability, use a 2nd person to work with the original researcher. They could read the source (separately) and devise a set of categories. They could tally the occurrences of each of the categories of the source and then compare tally charts looking for agreements. Researchers could calculate correlation between the two ratings.
What is meant by the significance at p>0.05?
It is the different that has a 1 in 20 probability of being caused by chance.
- less that 5% probability that the number/difference in whatever it is is left to chance
What does it mean for a test to have high concurrent validity?
When it compares a new test to an existing test (of the same nature) to see if they produce similar results. Scores can be correlated and a 0.8 positive correlation is considered internally valid. If this is the case, the new test is said to have high concurrent validity.
What’s one strength of time sampling?
Gather a broader variety of observations over a longer period of time, e.g. observing for 15 mins over a 30 minute period rather than 30 minutes straight - observations may not be as reliable due to exhaustion.
What’s one weakness of time sampling?
May miss some behaviours when they are having a break from observing.- so unrepresentative
How does repeated measures help an experiment?
- avoids ptp variables or individual differences in ptp DV as they take park in both conditions
- also measures the difference of the own ptps improvements
What does it mean if the results in a study were significant at the 0.01 level?
There is only 1% possibility that the difference in results between the 2 conditions is due to chance
What’s a way a psychologist could reduce the chance of a type II error?
Increase the level of significance to 5% (0.05)
What is a type II error?
It is a false negative - concluded that there was no significant difference in results between the conditions
What are the benefits of peer review?
Reduce chances of flawed research being published and therefore treated as ‘fact’ and misinforming the public.
Improves quality of reports by suggesting amendments and further control
What is a problem with correlations?
Correlation does not mean causation - there is just a relationship. A 3rd factor could cause the DV (extraneous variables)
What is meant by the term meta-analysis?
When a researcher collects multiple previously conducted different studies on the same topic and they are statistically tested and combines it to provide an overall conclusion
How can you do matched pairs?
Rank order results of the DV and the top two names on the list can both be written on separate pieces of paper and put into a hat. First name gets put into one group (specify the group names) and the second name goes into the other group. Keep putting the next two names into the hat separately and repeated this for all the ptps. This allows for the groups to be almost equal and therefore much more valid when comparing.
