Part 2: Research Methods

The research methods in applied social psychology are very similar to that of other fields of psychology that you may have studied. If you have taken PSYCH 301, much of this will be review for you. Thinking back to the scientific process, this section will focus on the hypothesis testing side of things.

Methods of Data Collection

One widely used method for collecting data in social psychology is to rely on self-report, that is, we ask participants questions and allow them to report to us what they are think, how they feel, or how they behave. Most commonly, this is done via surveys. As you will read in Chapter 3, there are several different types of survey-style questions that can be utilized, which I won't repeat here. One thing to really consider is how the wording of the survey questions can impact the results. Researchers must be sure that their questions are worded precisely, in a way that all respondents will understand, and do not lead participants towards a given choice (i.e., "loaded questions"). Additionally, the order that questions are persented in should be taken into consideration. For instance, it is highly recommended not to ask participants information about their demographics (e.g., age, gender, race/ethnicity) at the beginning of a survey, as 1) participants may begin to become bored by the survey and start answering more rapidly/less honestly toward the end simply to finish, or 2) if the survey is assessing differences as a function of one of those demographic questions, the participant may be tipped off about the purpose of the study and may alter their answers in response to this. 

Survey questions can either be closed-ended (participants are given choices for responses) or open-ended (participants can type/write in their own responses). By far, most surveys utilize closed-ended options, as these are far easier to analyze when examining the data. If we ask participants to report their level of anxiety on a scale of 1 to 7, we can plug that number directly into statistical software to run our analyses. But what if we ask participants to describe their level of anxiety? How do we compare their anxiety to that of other participants? Not such a simple task at this point. Open-ended responses are popular in qualtiative research (more on this later), and they are sometimes used in demographics questions. For instance, they can be used to ask for a participant's gender, allowing for gender nonbinary participants to respond in a way that most accurately reflects their social identity.

Survey research has exploded with the invention of the internet. Never has data been so readily available to social psychologists. Within minutes, data can be collected from participants all across the planet. With that said, there are some caveats to this type of research. For one, the data seems to be lower quality than data collected in-person (Heerwegh & Loosveltd, 2008). Participants may be more likely to respond at random, especially without someone watching over them. Another concern is whether there are real people answering your questions. Amazon MTurk is a platform that is used by many researchers to get access to cheap data, where participants can be paid for pennies on the dollar. However, there have been reports of individuals using bots to automatically respond to surveys, allowing them to earn money for participation without actually filling out the surveys, rending the data collected unusable. Another concern, which impacts many types of survey research, are lower response rates. Most survey requests that are sent out are ignored, thus only a small subset of indviduals will actually complete the surveys. This may bias the sample of respondents, for instance, perhaps only participants who have a lot of free time and are high in the personality trait of conscientiousness are willing to fill out the surveys. Thus the results may not generalize to other populations of individuals.

Another general limitation of surveys and other self-report measures....participants may not be completely honest in their responses, or they may lack self-awareness. For instance, a survey designed to assess how racist a participant is will probably not elicit completely honest responses from all participants. Participants may respond in socially desirable ways, that is, they may respond in ways that make themselves look better. To get around some of these limitations, researchers may use observational methods.

Observational methods involve directly observing the participants' behaviors. For instance, if we are interested in how a participant will act toward a member of a different race, rather than simply ask survey questions that they may not feel that they can answer honestly, we can set up a scenario where they are forced to interact with a member of a different race in the lab, and observe how they behave. For instance, their facial expressions could be recorded and coded for how much they smile or scowl in the cross-race interaction. The chapter will get into the different types of observational methods involved. It is important to note that these methods tend to be more time and resource intensive, hence surveys are more popularly used, especially in applied social psychology. However, when possible, observational methods are often superior.