Why we rarely use the word “prove” in psychology…

Travis DixonAssessment (IB), Revision and Exam Preparation

Your writing shows your thinking, so choose your words carefully.

If you’re reading this it’s probably because your teacher has sent you here after you’ve gone and written something like, “this study proves that…” in your psychology writing. While you may be able to use the word “prove” in the natural sciences and other subjects, this post will help you to understand why as a psychologist you may never use this word again.

The word prove is definitive, it’s absolute. It means that something has been demonstrated beyond doubt. When you’ve used this in your writing it sounds like you’re saying, “this study shows beyond doubt that…”

But in psychology there’s rarely a conclusion in research that is completely without doubt. In fact, throughout this course you should be practicing your ability to find aspects of a study to raise doubts – this is at the heart of critical thinking. To say that a study proves something is to reveal your lack of critical thinking, even in a short answer response.

When you go and say “this study proves that…” it suggests that you have a limited understanding of psychological research, and the complexities of human behaviour. 

Let’s look at some examples…

Every year I will have one student that says…“Loftus and Palmer’s research on leading questions proves that memory is unreliable and we can have false memories implanted in our minds.” 

Here are some reasons why this study doesn’t “prove” anything:

  • Only 32% of participants in the smashed condition say they saw broken glass in the video of the car crash. This means that 68% of participants memory was not altered and may have been reliable. This raises doubt.
  • This study was conducted in 1974, in the United States on college-aged participants. We cannot know for sure that these results would apply today, on other age groups and/or in different cultures. This raises doubt. 
  • The experiment was conducted in controlled circumstances and arguably does not reflect a real life situation. Could we really expect people’s memory to be unreliable in other situations? This raises doubt.
  • This study also only looks at explicit memory and in a very particular situation – this study can’t “prove” anything about implicit memory. This raises doubt.
  • We do not know for sure that they actually remembered seeing broken glass. They checked “yes” on the question sheet, but this doesn’t actually mean they remembered the broken glass? Maybe they were being lazy, didn’t really think carefully or perhaps they were demonstrating participant expectancy effects. This raises doubt.

All these issues raise at least some doubt about this study’s ability to “prove” that human memory is unreliable and can be manipulated. And one little doubt is all it takes for the word “prove” to become false and erroneous when explaining conclusions to research. If there’s just a hint of a doubt, the word “prove” should not be used. So when you use the word prove, what you’re declaring is: “I can’t see anything wrong with this perfect study and we’d get the same results across all people, at all times, in all places.” This is rarely the case in psychological research because humans are so different.

Single Studies Can’t “Prove” Anything

Another reason why your use of “prove” is probably wrong is because a single study can’t prove much. Before we cant say we’ve proven anything, the results have to be replicated many times. For example, we can now say that it’s been proven that our brains change as a result of experience. This is an accurate experience. However, this didn’t just happen after one study – I can only say this confidently because this has been shown time and again in hundreds (if not thousands) or studies.

To use the word “prove” incorrectly when writing about research in an exam answer is like dribbling your tomato soup right down your new white shirt while on a date: what you did up until that moment has lost it’s charm and been overshadowed by your sloppy mistake, and everything you do after will always be tainted by that big red splurge of a reminder right in the middle of your page. 

Let’s look at another example:

HM’s case study proves that the hippocampus is responsible for transferring short-term memory into long term memory.” 

Remember, all we need is just one detail that might cast a tiny hint of doubt on whether or not these results could apply to all humans, at all times, in all places. We needn’t look much further than the fact that HM is just one man. Can we use the results from just one person to make such a bold, definitive and absolute claim about a part of the brain and memory for all humans? There is at least some doubt. To use the word “prove” here would show you haven’t thought carefully about the study and/or you don’t understand the concept of generalizability. Play it safe and use a softer verb.

So what can we say instead?

Here are some better verbs to use in place of “prove”:

  • suggests
  • demonstrates
  • reveals
  • indicates
  • shows
  • confirms

I know what you’re thinking – “They mean the same thing!” They’re close, but not identical. The word “prove” is much stronger than these words. As mentioned earlier, it doesn’t leave room for doubt or uncertainty. And if there’s even a hint of doubt (as there nearly always will be in psychological research) we should not use the word prove.

Here are some example conclusions to research using these alternative verbs:

  • Asch’s research on conformity shows how individual behaviour may be influenced by group norms.
  • Bandura’s Bobo doll experiment demonstrates the process of observational learning and how children might learn to copy aggressive actions by observing others, especially if they can relate with with the model.
  • Milgram’s experimental results on obedience suggest that people may be inclined to inflict suffering on others if they are following orders by someone in a position of authority.

With a range of alternative verbs at your disposal, there really is no need to use the word “prove” and run the risk of proving your lack of thinking and understanding.

The importance of qualifiers

Along with the absence of the word “prove” in the above examples, notice how these conclusions also use qualifiers, such as “might” and “may.” These also help to convey the understanding that the results of the studies are not definitive, they are not absolutes, and there may be uncertainty generalizing these findings to other situations, people, or places.

Let’s look at two answers for comparison that come from IB Psychology: A Student’s Guide… 

  • In one of her answers, Sarah writes that “…because of their cultural values, Japanese people go along with the group.”
  • In one of his answers, Raffi writes that “…because of their cultural values, Japanese people may tend to go along with the group.”

I really want to emphasize the importance of the difference in these two statements and why Raffi’s is far more accurate and demonstrates a deeper understanding than Sarah’s.

  • Sarah is making a definitive claim about Japanese people that suggests she thinks of all Japanese people being exactly the same and behaving the same way when in a group. Sarah is treating all 120 million + Japanese people as the same, which means she either hasn’t thought carefully about the conclusion she is drawing, or she hasn’t thought about the words she is using to show her conclusions (Surely she doesn’t really think they all behave the same?)
  • Raffi, on the other hand, includes two very important qualifiers to his claim: “may” and “tend to”. The word “may” is important in this conclusion because it recognizes that Japanese people won’t behave the same way all the time. It also allows for differences in individuals, as not all Japanese people may share the same cultural values. The phrase “tend to” also shows that while this is a pattern of behaviour, it leaves it open for exceptions. Be life Raffi!

My students often make fun of me for always reminding them of qualifiers and I think they call me something like “Mister maybe suggests that this might possibly happen if…” behind my back. I’m OK with this; I’ve been called worse. But I hope you can see from these examples why making oversimplified claims in your conclusions can really show a lack of thinking and understanding in your writing.

Remember: language conveys meaning, so choose your words carefully. 


Has a psychologist ever “proven” anything?

There are some instances when research has proven certain things to be true. But this is not often the case for a single study. It is more likely that a single study’s findings led to further research and it’s only after much corroboration and successful replications that anything is proven. It is more likely that a psychologist’s study has helped to prove something.

For example, Eric Kandel has been conducting research using sea snails for decades and his research has helped to prove that memory storage has a neurological base. Notice, however, that this is a very broad concept that his research has helped to prove. And this is not the result of a single study – it has taken decades of research to get to this conclusion.

One could argue that Loftus’ research has helped to prove that eye-witness testimony may not be reliable. But notice even in this statement I must say “may not be reliable.” If I was to change this slightly and say, she’s “helped prove that eye-witness testimony is not reliable” I am back to making an overly definitive and absolute statement that is saying categorically and without a doubt that no eye-witness testimony could ever be trusted because it’s unreliable. This simply is not true.

But even with these final examples I’m hesitant to use the word “prove.” And why would I, when I can use the word “demonstrate” or “suggest” just as easily? These qualifying verbs are still helping my explanation of what the research has shown, while still allowing for some form of doubt and questioning.

Language conveys meaning and your writing is communicating your understanding. For this reason you must be very mindful of the words you are using so that your full understanding of psychological research is conveyed accurately to your reader. 

A tip for teachers (and a warning for students)

I don’t ban the use of the word “prove” in my classroom; I know students will use it in the first semester, as they’re still learning the subject of psychology and its quirks and kinks. If I was to make a blanket ban on the use of the word, I’d fear that I was running the risk of having my students graduate without ever learning why the term is rarely applicable in psychological research; I would be sacrificing compliance for understanding, and this I think is to borrow from Peter to pay Paul.

So when a student of mine uses the word “prove” when writing the conclusion of a study I set them an additional homework task: they are to find at least three factors that cast some doubt on the conclusion that the student has stated. In a nutshell, I get them to critique the study’s internal and/or external validity. I might even give them some guidance or get them to relate to something they know. For example, to use Loftus and Palmer’s example earlier in the post – I might ask this student, “How do we know that not all participants had false memories?” Or when looking at other studies I might ask them (as I teach mainly Japanese students), “What’s one reason why these results might not be the same for Japanese people?”

And from now, I’ll get them to start by reading this post.

I’ve found that by getting students to challenge generalizability in response to the use of the word “prove” in answers, I can develop their critical thinking and hopefully ensure that they understand why it’s a poor word choice in psychology. This critical reflection I hope has a deeper impact than blindly following rules.

I hope this has been helpful.