Skepticism about social psychology

Most psychological research doesn't replicate, but social psych research is particularly bad.

There is no deep truth to be found in social psychology because society changes.
Social psychology does not discover any scientific theories or laws that have any chance of standing the test of time. No theory or law is possible in social psych. Everything social psych learns will be wrong within 30–100 years, just as the social research from the past is wrong today. A great deal of social research is already dubious at time of publication because it fails to take cross-cultural differences into account. The most interesting social psych studies from the past would be completely unethical today (e.g. Milgram Shock Experiments) and, worse yet, some were falsified or otherwise deeply problematic at their own time (e.g. Stanford Prison Experiment).

Social psychology can still be quite interesting.

Social psychology involves building a career on what is hot and topical right now.
This could be quite interesting to the person doing the research or to a public that is curious about current affairs.

For example, I went to a social psychology talk where a PhD student had analyzed the songs people said they listened to after a romantic breakup. This PhD student used the song lyrics to sort the songs into categories based on attachment style. This PhD student won an award for this project. That might sound like an interesting project that aligns with a curious mind, even though it would not be scientifically valuable across time.

Why not?

Do you think this research would replicate?
In principle, this research about breakup songs cannot translate back in time since contemporary songs hadn't been written yet: a sad person in the 1940s cannot listen to a song that came out in 2015. This research doesn't seem likely to translate forward in time, either: a sad person in the 2140s almost certainly won't listen to songs that came out in 2015 when they go through a breakup. Indeed, people today don't listen to songs from the 1840s when they go through breakups. No, this sort of research is not destined to replicate.

Enduring Findings

To my awareness, the enduring findings in social psychology can be summarized as follows:

• In general, groups tend to compete over resources, especially scarce resources, though this can be mitigated by cooperating on superordinate goals (e.g. disaster relief).
• In general, people tend to favour members of their in-group over members of out-groups.
• In general, people tend to express less explicit prejudice under conditions of cooperation for common goals in a supportive environment of their peers.
• In general, people tend to assume that other people act according to their personality (stable) rather than their situation or mood (transient), though this may be less applicable in collectivist cultures.
• Sometimes, having more strangers around can result in less individual helping behaviour (the Bystander Effect), though this doesn't necessarily apply in situations of unambiguous danger and does not apply to people that already have a relationship.

Social psychology has a dubious history

Most of what I learned that went by the name social psychology turned out to be very flashy studies from decades long ago: the Stanford Prison Experiment, the Milgram shock experiments, the Asch conformity research. Years after their publication, these studies turned out to be quite dubious. They were often either corrupt and unethical from the very beginning, deceptively misreported, or not replicated in any recent history.

Since transitioning to graduate school, a fair amount of the research I've seen that has been called social psychology has been sentiment-analysis of Twitter/X threads about trendy topics, which are not even in principle scientifically generalizable cross-culturally or to future or past human societies.

If you are deeply interested in relatively transient social phenomena, especially emerging online phenomena, and are not particularly committed to long-term scientific discovery, social psychology might align with your interests. Social psychology allows you to collect data about trendy topics. Personally, I can't help but conceptualize this aspect of social psychology as providing more of a journalistic or historiographical approach to knowledge and current affairs, edging closer to humanities than the natural sciences.

Personally, any interest I had in social psychology was discouraged by its lack of long-term scientific impact. However, I found new insights by studying Game Theory, which is a much more precise mathematical study of strategic interactions. Game Theory can account for a number of social phenomena with precise mathematical models. If you are interested in social phenomena, check out Game Theory to see if it satisfies your curiosity.

Social psychology cannot account for individual differences

Human beings are not fungible: one cannot remove one person and slot in another, expecting the two to be equivalent.

Consider how the Milgram shock experiments are often misunderstood.
The ostensible conclusion drawn from these studies is that "X% of people would shock strangers to death if told to do so by a person in a lab-coat".

That isn't quite right, though. You cannot pick a person randomly and know what they will do. Each individual person doesn't have X% chance of being the kind of person that would shock someone to death.

There are sub-groups. Generally speaking:

• (i) some people refuse to administer any shocks,
• (ii) some people administer shocks until the (confederate) participant says to stop,
• (iii) some people keep going after the (confederate) participant says to stop. The people in group (iii) tended to persist in shocking all the way to the end.

This subtler understanding of sub-groups offers very different interpretations!
Instead of thinking that anyone might shock someone to death, we get a very different picture of three types of people:

• (i) people that have a "do no harm" principle,
• (ii) people that focus on consent,
• (iii) people that focus on authority.

To understand the mediators and moderators of this population further, we would need to separate them into sub-groups, not keep acting like they are all the same single group with an X% chance of being in group-(iii).

Index

Return to Psychology The State of the Field

Continue to Skepticism about evolutionary psychology

Jump to Undergraduate Psychology