"Research Methodology" Close scientific method based on empi

LECTURER: Good morning. Now, what is it you want to discuss today?
STUDENT: Good morning, Dr Reed. This assignment you’ve given us is the first psychology experiment I’ve had to do and I’m not sure where to begin or which steps to take.
LECTURER: Well, conducting your first psychology experiment can be quite a complicated and confusing process but just remember that like other sciences, psychology uses the scientific method and bases its conclusions upon empirical evidence.
STUDENT: What do you mean by "empirical" evidence?
LECTURER: Ah, well, empirical evidence is established by observation rather than theory.
STUDENT: And the scientific method?
LECTURER: Oh, yes, when conducting an experiment you need to follow a few basic steps.
STUDENT: I know the first step is to come up with a research question or problem.
LECTURER: Yes, a question that can be tested.
STUDENT: How do I find an appropriate question?
LECTURER: I would suggest one of three methods. Firstly, you can investigate a commonly held belief or what we call "folk" psychology.
STUDENT: I see. So, I could examine the belief that staying up all night to study for an important exam can adversely affect test performance?
LECTURER: That’s right. In that case you would compare the scores of students who stayed up all night with those of students who got a good night’s sleep.
STUDENT: I think I could do that.
LECTURER: Well, alternatively, you might want to consider reviewing the literature on psychology. You know, published studies can be a good source of unanswered research questions. I’m sure you’ve read papers where the authors note the need for further research.
STUDENT: So I would come up with some questions that remain unanswered?
LECTURER: Correct. But there is a third source of ideas: just think about everyday problems and then consider how you could investigate potential solutions.
STUDENT: Okay...perhaps I could study various memorization strategies to find out which are the most effective.
LECTURER: That’s the idea. Next, you need to define the variables--you know, anything that might have an effect on the outcome of your research.
STUDENT: Yes, I remember we learnt about that last week.
LECTURER: Yes. That’s right. Then you have to develop a testable hypothesis that predicts how the variables are related.
STUDENT: For example: "Students who are sleep deprived will perform worse in an exam than students who are not sleep deprived"?
LECTURER: Exactly. Once you have developed a hypothesis you must carry out background research.
STUDENT: I can use books, journals, online databases and websites?
LECTURER: Yes, all of those. I covered the reasons for background research in last Friday’s lecture, didn’t I? What you have to remember at this stage is to take careful notes and generate a bibliography of your sources.
STUDENT: Okay, I’ve got that...then, I’m ready to develop an experimental design?
LECTURER: Well, again you have a choice. There are three basic designs and each has its own strengths and weaknesses. The pre-experimental design does not include a control group so there is no comparison. What we call a quasi-experimental design does incorporate a control group but there is no randomization whereas a true experimental design has both control groups and random assignment to groups.
STUDENT: You’ve also told us about standardization of procedures—is this where that comes in?—being sure to compare apples to apples...
LECTURER: Absolutely. Going back to your sleep deprivation example...the same exam would have to be given to each participant in the same way at the same time, etcetera.
STUDENT: Got it.
LECTURER: When selecting subjects, you need to consider different techniques. If you were to go through with your sleep deprivation experiment, you would need to ensure that your experimental and control groups were standardized, that is, all third year accounting students, for instance. A simple random sample involves choosing a number of participants from a group of similar people. On the other hand, a different kind of study might involve a stratified random sample where participants are randomly chosen from different subsets of the population.
STUDENT: You mean...subsets with distinctive characteristics...like age, gender, race, socioeconomic status and so on.
LECTURER: Precisely. Then the next step is to actually conduct the experiment and collect the data.
STUDENT: Then I have to analyze the data.
LECTURER: I’ll be dealing with the statistical methods for analyzing data in next week’s lecture.
STUDENT: Oh good. I guess all that’s left then is to write up the data?
LECTURER: Yes, communicating your results is important and in the next couple of lectures I’ll be covering the format and structure of a psychology paper and tips for writing each section.
STUDENT: Thank you, Dr Reed. I feel much more confident in getting started now. Thank you for taking the time to see me.