If you want to write excellent evaluations of psychological studies then generalizability is a really important term to know. It’s also called external validity and it refers to the extent to which we could expect the same results in a different context (i.e. do the results apply beyond the study, external to the study?).
In this post we’ll focus on population validity (read more here). Strictly speaking, population validity refers to how well the sample represents the target population – is the sample an accurate representation so that we’d be able to generalize our results to the broader group. It can also be used to assess the extent to which results from one sample could apply to a different group of people.
When you’re evaluating a study based on generalizability what you’re really doing is evaluating the conclusions of that study. So for population validity, we’re asking “would those conclusions apply to everyone? Are there specific groups they might not apply to and why not?”
If a study is done on all women, we might ask if we’d expect the same results with men. This is an example of thinking about generalizability. (Photo by rawpixel.com on Pexels.com)
For example: “If one study used all females and concluded that there’s a relationship between testosterone and aggression, could we apply these findings to males as well? Males naturally have higher levels of testosterone and perhaps this has led to other biological differences that might affect the processing of testosterone in the body. Moreover, if the findings of the study were being applied to explain high testosterone levels in prison populations, the all-female participants also raises the question of population validity as a majority of violent criminals (statistically speaking) are males. Thus, the nature of the sample in the study raises questions about the generalizability of the possible applications.” (From IB Psychology: A STudent’s Guide pg35).
Activity
Read the following list of studies with a brief description of the sample (i.e. participant characteristics) and the conclusions drawn from these studies. Discuss limitations of these studies in terms of population validity. The key to a good explanation is to pick a group or population that you think the results wouldn’t apply to and explain why not!
Testosterone and the amygdala (Radke et al): This study used 54 healthy females with an average age of 21 (range of 18-30 to test the effects of testosterone on the amygdala when someone is threatened. The results showed increased testosterone increased activity in the amygdala when responding to threat. One conclusion from this study is that testosterone may cause aggression because it increases activity in the amygdala.
- Population validity points to consider: gender, age, health.
Damage to the brain and aggression (Grafman et al.): In this study, Vietnam War Veterans (mostly males) who had damage to the brain were compared with other Veterans who had no damage. The results showed that damage to the vmPFC in particular was associated with more violence. One conclusion from this study could be that damage to the vmPFC causes aggression.
Most studies are conducted in America and on Americans. If you are saying that you think the results might not apply to a different culture or country, try to be specific and give a reason why you think they might not. (Photo by Todd Trapani on Pexels.com)
- Population validity points to consider: Nationality (American), age, occupation.
Culture of Honor and Aggression (Cohen et al.): These studies used white, American, college-aged, male participants from the Northern and Southern states (along the East Coast of America). The results showed that Southerners were more likely to be primed (ready) to respond to a threat with aggression. One conclusion from this study is that people from a Culture of Honor are more likely to respond with aggression or violence to a threat.
- Population validity points to consider: Nationality (American), age, race, occupation.
Bobo Doll Experiments (Bandura, 1961/63): These studies were conducted on 3-6 year olds who were enrolled at the Stanford University Nursery in Stanford, California, USA. The kids watched an adult (model) behave violently with a doll or nicely with it and then the researchers measured how many aggressive acts each kid performed when left alone with the toys. The results showed that watching an aggressive model increased aggression in the children. One conclusion from this study is that violence can be a learned behaviour and we may learn it through the observation of others.
- Population validity points to consider: Nationality (American), age, race (mostly white).
Exam Tip: If you’re not sure which type of validity you’re writing about, just refer to “generalizability” and explain how or why you think the results might not be applicable to another group of people, situation or setting. The best explanations will be 3-5 sentences and will give reasons why we might not expect the same results.
Travis Dixon is an IB Psychology teacher, author, workshop leader, examiner and IA moderator.