1.3 Research Methods: Empiricism, Hypothesizing, and Experimental Design

Most research in psychology is conducted in universities by faculty members and students, so many participants come from introductory psychology courses or the main student body. Undergraduate college students taking an introduction to psychology course are not a good representation of the human population. Not only is a large portion of psychological research conducted on undergraduate college students, but participants are also typically overwhelmingly from Western, Educated, Industrialized, Rich, and Democratic societies, or what has come to be called the WEIRD problem A problem in psychology where research participants are not a good representation of the whole human population, as the participants are often from western, educated, industrialized, rich, and democratic societies. in psychological research (Prayogshala, 2020). These problems create potential biases in the research and a tendency to overgeneralize the results to people from different cultures, socioeconomic statuses, education levels, and experiences.

Arnett (2009) evaluated articles from different subdisciplines in American Psychological Association (APA) journals and found 70 percent of participants were from the United States, and 96 percent were from WEIRD countries (refer to “The Hidden Biases in WEIRD Psychology Research”). Recently, Arnett and colleagues revaluated APA journals and found a slight improvement, with about 60 percent (a drop of 10 percent) of participants coming from the United States and 89 percent coming from WEIRD countries (Thalmayer, Toscanelli, & Arnett, 2020). Understanding these deficits is difficult because this demographic information has not been traditionally assessed or reported in psychological journals (Rad, Martingano, & Gingers, 2018).

There are concerted efforts to broaden the scope and diversity of participants recruited for psychological studies and to increase the diversity of those conducting the research. Many journals require authors to publish demographic information about their participants even when this information is not a specific variable in their design. Rad et al. (2018) made several suggestions for authors and publishers to reduce the WEIRD problem in psychological research:

Using human subjects in experiments has a troubled history of exploitation, racism, and cruelty. In the early part of the twentieth century, no rules governed the use of human subjects in medical or psychological research. In the 1940s, the Nazis conducted horrific and torturous medical experiments on those held in concentration camps. This led to the Nuremberg Code, established in 1948, which is an international document requiring voluntary and informed consent by human participation in research (Komesaroff et al., 2002). From 1932 to 1972, the Public Health Service in the United States conducted a study on the effects of syphilis on Black men in the infamous Tuskegee Syphilis Study, in which the men were misinformed about the purpose of the study and were not given proper treatment for syphilis even when a treatment (penicillin) became available in the 1950s. The Tuskegee Syphilis Study led to a National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (NCPHS). In 1979, the NCPHS created the Belmont Report (Department of Health, Education, and Welfare, 1979) that laid out three basic principles for conducting ethical research with human participants:

The Belmont Report informed the ethical treatment of human participants in psychological research. Guidelines and principles for working with human subjects are typically very similar across universities. When planning a research project involving humans, a proposal is written containing detailed information, a literature review and the purpose of the study, and a description of the methods and the informed consent form, so participants know the parameters of the experiments and their rights. The proposal also provides the committee with details about the intended participants and how they might benefit from being involved in the experiment (e.g., money, course credit, education) as well as possible risks to their physical or psychological well-being. An internal review board (IRB) A group of colleagues and representatives from the community that evaluates whether the proposed research will treat participants ethically. evaluates the proposal and determines if the experiment can proceed, if more information is needed in the proposal, or if it is rejected. There has been psychological research carried out in the past that would likely not pass an IRB evaluation today (refer to “Five Psychology Experiments You Couldn't Do Today”).

Animals are used in several lines of research in psychology. The foundation for using animals in psychological research is the strong evidence that there is continuity in the fundamental processes underlying behavior in many animals, including humans.

Today, rats and pigeons make up approximately 90 percent of animals used in psychological research (APA, 2010). Pigeons are used in research on learning and cognition because they see color and are highly motivated to receive food rewards (refer to Figure 1.12). Rats and mice are often used in research because they live for two years (which allows for aging studies), are relatively inexpensive to keep and maintain, are easy to keep in a limited space, are easily motivated by food, and learn quickly. Another recent advantage of using rodents in research is the development of techniques to alter genetics through inbreeding, crossbreeding, or changing or removing selected genes through genetic engineering.

Strict guidelines and oversight protect animals that are used in biomedical and psychological research. The Animal Welfare Act, created in 1996, established rules and regulations for the ethical treatment of animals. Animal use is overseen by the United States Department of Agriculture (USDA) and the Animal and Plant Health Inspection Service (APHIS). These organizations set strict guidelines for how animals are housed, fed, used in research, and euthanized. There are also strict guidelines about how and why research is performed with animals. 

A correlation is when two or more measurements change together in predictable ways. In a correlational research design A research design where the experimenter looks for how data collected about different variables change or do not change together. , the researcher does not have to manipulate the experience of the participants by changing levels of an independent variable (though they can); they simply collect data and see how different variables change or don’t change together. Two variables can increase or decrease together, which is called a positive correlation, or as one variable increases, the other can decrease, which is a negative correlation.

Let’s take the example of our second hypothesis in our sleep study. Suppose we give a survey to one hundred students and ask several questions about their sleep habits as well as their school performance, and we find that the average number of hours of sleep per night shows a positive correlation with their cumulative GPA—as sleep hours increase, GPA increases. Scatter plots A graphing design where a single point represents an individual’s response on two variables presented on the x- and y-axis of the graph—often used to demonstrate correlations. are typically used to showcase correlational data, with the horizontal x-axis indicating one variable (e.g., average hours slept per night) and the vertical y-axis indicating the other variable (e.g., GPA) (refer to A in Figure 1.13). We could also look at another aspect of sleep, such as the number of times a person wakes up during the night compared with GPA, and find a negative correlation (refer to B in Figure 1.13). Finally, we see no correlation between the time someone gets up in the morning and GPA (refer to C in Figure 1.13).

We need to be careful not to fall into the problem of correlation automatically implying causation. Looking at A in Figure 1.13, one might assume that increased sleep caused GPA to increase. While sleep might be a contributing factor in getting better grades, there may also be other factors that cause the increase in both. Can you think of a factor that might affect both sleep and grades? What about the level of stress a person is experiencing? Someone experiencing a stressful life may have trouble sleeping and have difficulty concentrating in class and studying. One of the goals of correlation research is to try to understand better why variables change together. 

A longitudinal research design An experimental design that involves recording the same measurements over a short or long period of time, often from the same participants. involves recording the same measurements over a short or long period of time. Developmental psychologists often use longitudinal research to track how age, time, and experiences change measurements of memory, life satisfaction, drug use, reasoning about morality, sleep patterns, sexual behavior, and many other psychological factors. For example, the Minnesota Twin Study has been surveying identical and fraternal twins for decades to better understand the genetic and environmental influence on psychological measures (Segal, 2012). Longitudinal data are often presented as line graphs with time along the horizontal x-axis and the dependent variable on the vertical y-axis. For example, suppose a researcher is interested in how someone would score on a sensation-seeking survey as they age. The survey might ask questions about how much someone might enjoy roller-coaster riding, sky diving, and cliff jumping if they were given the opportunity (refer to Figure 1.14).

A longitudinal research design is a powerful way of tracking changes at different ages, but it can be expensive, and because it may take several years, researchers run the risk of losing contact with participants along the way. Another way to investigate age-related differences is to conduct a cross-sectional design An experimental design that samples from a diverse population, such as different age groups. in which a sample of the population at different age groups is evaluated and compared. For example, suppose a researcher is interested in how the prevalence of drug addiction changes from adolescence to middle adulthood. They might provide a survey that evaluates problematic drug use to fifty people in different age groups. A histogram or bar graph might be used to display the results (refer to Figure 1.15).

A naturalistic observation A research method that takes a hands-off approach to data collection, such as watching children interact in a classroom. method of research takes a hands-off approach to data collection. The researcher puts themselves in a position where they observe the group of people, watch videos, and take notes, but do not interfere or interact with subjects of the research. This could happen in a laboratory by watching people through a one-way mirror, but it is often carried out in the field by observing children in a classroom, shoppers at a grocery store, or employees working with equipment in a factory, for instance. Naturalistic observation techniques are used in several other fields, such as anthropology and zoology. Jane Goodall moved to the Gombe Reserve in Africa in 1957 and observed chimpanzees in the wild for several years, where she discovered many unique things about these apes, including that they used tools.

Naturalistic observation lacks the ability to control participants’ experiences and can run the risk of subjective interpretation of what is observed, but it has the advantage of conducting research in more realistic and complex environments. One way to help reduce subjectivity is for researchers to watch videos and score what they see using an agreed upon rubric describing specific behaviors. 

Research in psychology often involves evaluating many participants, which allows for greater generalization of the findings to other members of the population. However, there are times we can learn a great deal from single individuals with unique problems, circumstances, or abilities. An in-depth investigation of one person or situation is known as a case study A method where research is conducted with one participant who has unique characteristics. . A famous case study we will learn more about in Chapter 7 “Motivation and Emotions” is that of Henry Molaison (known as H. M.), who had a large section of his brain removed when he was a child to prevent severe epileptic seizures. After his surgery, H. M. was not able to make new long-term memories. As one would imagine, a researcher could not replicate these situations in the controlled setting of a laboratory, but case studies can provide insights into neuroscience, memory, and psychological disorders by looking at real-world examples that cannot be produced in a lab. 

Several large research projects collect a great deal of data that they make available to other researchers and, in some cases, the general public. This archival data A method of conducting research by evaluating data that has already been collected and made available to the researcher. set often has demographic information about sex, age, ethnicity, income, political affiliation, religion, and location. It might also include results of survey questions on topics related to psychology, such as stress levels, childhood experiences, or life satisfaction. Research with specific questions can then use the archival data instead of collecting their own to test hypotheses. This APA webpage has links to several archival data sets on topics ranging from child language to adolescent health.

Like archival data research, a meta-analysisA method of research where results from many studies that investigate the same or similar topic are reviewed and summarized. It is good for looking at larger trends. does not develop and run new experiments but instead carefully evaluates previously acquired data. In a meta-analysis, researchers use statistics to evaluate the results of multiple scientific studies investigating the same question. This approach has the advantage of not relying on the results of a single study for answers, as it looks at larger trends by evaluating the results of many studies. Scientists often write systematic reviews of a specific topic where they synthesize the findings of many decades of research, and they often use a meta-analysis in forming conclusions. A colleague of mine, Dr. Chris Miller, and his students conduct meta-analyses of fMRI brain imaging studies that are interested in psychological disorders. For example, in a recent line of research, Dr. Miller and his team searched for all imaging studies that have looked at depression, and then they created computer models that evaluate patterns of brain activity across the images from all the studies. In later sections, we see how meta-analysis works well to combat several of the problems in psychological research (for review, refer to Stone & Rosopa, 2017).