Improvements of studies Flashcards
How to improve L&P’s study ecological validity?
I would use a staged car crash outside using real cars driven by stuntmen, rather than showing P’s car crash videos.
This is an improvement because P’s reactions will be more true to life as seeing real cars crash will simulate the stress of seeing a real car crash.
However, the change will make the study less ethical as the stress caused by the car crash might causes P’s to experience psychological distress.
How to improve L&P’s generalizability?
I would use volunteer sampling and put up an advert for the study in a location such as a shopping center where many ages of people will see and sign up. This will be done instead of an opportunity sample of students.
The sample will produce findings which are relevant to both drivers and non-drivers, whereas in the original study, the age bias towards students would have meant that very few were driver.
However the change will make it slower to conduct the study as the researcher will need to wait for P’s to sign up in response to the advert.
How to improve Grant’s ecological validity?
I would give P’s 2 hours of break before taking the Psychoimmunology test, rather than letting them only have a 2-minute break.
P’s performance in all conditions will be truer to life. This is because as the breaktime is shorter, a large part of the information from the article would still remain dormant in the P’s minds, their memory has not yet been distorted while in real life, there tends to be a longer gap between studying and taking the test.
However, the change will make it slower to conduct the study as the break between encoding and retrieval would be longer.
How improve Grant’s data type?
I would collect qualitative data by asking P’s to blurt all the information they remember about the article on a paper. As free recall has been used, researchers would gain a greater insight into which details were remembered and forgotten. Improving their understanding of thought processes.
However, the change will make it harder for researchers to use inferential tests to find out how context of the encoding and retrieval interact with each other.
