IA Tips for Loftus and Palmer

Travis DixonInternal Assessment (IB)

Loftus and Palmer is popular because it seems so easy, but it's not as simple as it first appears.

By far it’s the most popular study to conduct for the IB Psychology IA. But watch out, the “car crash study” can cause major accidents if you’re not careful. So if you’re doing Loftus and Palmer for the IA, read this post carefully. 

Below is a summary of some common errors students make and tips on how to avoid these errors. I’ve begun with the most basic and then I’ve finished with a real extra for experts.

Read More:


Two conditions only: The classic mistake in IAs is that students choose all five verbs – bumped, smashed, crashed, hit, and contacted. This makes the statistical analysis far more complicated than it needs to be. The easy solution is to choose two. Come up with a good reason for the two you have chosen and you can explain this in your materials (why you chose these verbs to put on your questionnaires).

How to measure the DV: For some peculiar reason, students often use a likert scale to gather their data. For instance, they will ask:

  • “How fast were the cars going when they (?) into each?” 
    • Circle One: 0-10kph, 11-20kph, 21-30kph, 31-40kph, >41kph

Or something similar to this. I would advise that you avoid using this likert scale and you simply leave a blank space for the response. Perhaps give some guidance like, “you can answer in miles per hour or kilometers per hour” so they give a speed estimate, and don’t say “Very fast” or something similar.

Save time and money with our step-by-step IA eBook and Teacher Support Pack bundle.

Likert scales are another way the data analysis could become tricky and you run the risk of using the wrong statistical tests because of how the IB treats likert scales (they are considered ordinal data in IB, so the median should be used).

Explain your materials – the video: When explaining your materials, focus first on those directly involved in manipulating and measuring your variables. In this case, the choice of video is an important material you’ve chosen. I’m sure your group watched at least a few before you chose which one to use, right? Then think about what was your selection criteria. What made you choose the video you used?  If you followed the advice above and chose a video with a gentle crash in which the cars don’t blatantly “smash” into each other, you have a good reason to explain why you chose this.

Bonus tip: the original video is not available online (at least it wasn’t at the time of writing). There are a lot of videos you can use, however. Just don’t say it was the original one.

Explain your list of questions: Most students ask a few questions in their questionnaire, not just the critical question regarding how fast the cars were travelling. Did you ask general memory questions like they did in the original? If so, why? Doing this can be a good way to control some variables, like reducing the demand characteristics of the procedures. It’s important to explain this in your exploration.

Stick to a car crash video: Some students like to alter the methods and do something a bit different to a car crash. For example, I had students one year film someone running and then they asked questions based on that film. Other groups have used different events, like dogs running, kids playing or people arguing. This is fine and if you’re really interested in doing this, go ahead. However, if you are going to change the scene then make sure you have a good reason. This adds another thing to explain in your materials. If you can’t think of a good reason, stick to a basic car crash. It’s easier and avoids problems.

Consider doing a different study: You won’t lose marks because this study is so popular, but there are other options. For example, you could do a study on the “Weapon Focus” – the phenomenon that occurs when people remember less details about an event when a weapon is present because their attention is diverted to the weapon. Of course, you don’t have to use a real weapon, but a photo with a weapon in it would suffice. Read more about Weapon Focus here.

The Big One!

The Background Theory

Choose your background theory/model carefully: There are a few options for which theory to use when explaining Loftus and Palmer. The two most popular are schema theory and reconstructive memory theory.

  • Reconstructive memory theory: Personally, I do not like this “theory” because I don’t actually think it’s a theory, per se. It’s difficult to find original journal articles or scholars identifying this theory by name (a search on google scholar yields just 7 results). The fact that our memories are not direct recollections but are more accurately described as reconstructions is a more general idea rather than a specific theory. This means using particular details of this “theory” are difficult so explaining the link to the investigation becomes tough (if someone can find a peer-reviewed article on “reconstructive memory theory” I’ll gladly change my opinion as it will make life much easier – just pop a comment down below).
  • Schema theory: I have found that explanations using schema theory are a bit hit and miss – some are good while others are weak. The key to using schema theory effectively is to use characteristics of schema to explain why the stronger verb produces a faster speed estimate. Top IAs reference journal articles to source the information they get about schemas that they can use to explain their effect on memory accuracy. This is a challenging task, but if you’re using schema theory as your background theory you should do this.

After a couple of years of examining the IA and seeing the importance of the background theory, I’d like to propose an alternative…

“The Strategic Effects Account” of the Misinformation Effect (Ayers and Reder, 1998)

The IB allows “effects” to count as models or theories, so in this case just the misinformation effect by itself should be enough. However, it’s even better if you can find a theory that explains the effect. EDIT: Since writing this the IB has said that “effects” cannot count as theories or models for the IA, so this info is now outdated.

  • Misinformation effect: The original study is a test of the misinformation effect, which is “…the inclusion of information that is inconsistent with the course of an event, and which originates from sources other than the event itself, into a witness’s report of the event” (Loftus and Hoffman, 1989, as cited in Szpitalak and Polczyk Link). However, one issue is that the first experiment doesn’t really give great evidence for this – that’s why L&P did the second experiment with the broken glass “yes/no” question (see this video on how to explain the study properly for more info). But it’s best not to do this as it gathers nominal data and that can make statistical analysis a bit more difficult. You can, however, still use the leading questions Experiment #1 and argue that an extreme verb, like “smashed” is a source of misinformation, especially if it’s unlikely to apply to the video you’re using. For example, if it’s a gentle crash and you use verb “smashed” to describe it in your leading question, it’s more obvious that this is false information. This false information causes a false memory of the crash, resulting in a higher speed estimate. (Other studies by Loftus are more direct manipulations of false information whereby they give factually incorrect details to later recall, like asking about a yield sign when it was a stop sign, but this is tricky to do experimentally for the IA).
  • “The Strategic Effects Account”:  Whereas L&P concluded in their research that the leading question directly alters the memory, McCloskly and Zarazoga argue in a “Strategic Effects Account” that the “…subjects sometimes simply forget the critical information that was presented, regardless of whether or not misinformation was also presented.” In such cases when they forget the information they use the false information to guide their answers. It’s not that there’s a false memory occurred, but rather the procedures of the task encourage the use of the false details to make an estimate rather than the information from the original video (which is forgotten) (McCloskly and Zarazoga, 1985, as cited in Ayers and Reder 1998). This is actually similar to one explanation that L&P gave for the results of their first experiment. They conceded that “…it is possible that the differential speed estimates result merely from response-bias factors. A subject is uncertain whether to say 30 mph or 40 mph, for example, and the verb smashed biases his response towards the higher estimate.” (Loftus and Palmer, 1974, p586). This theoretical explanation is arguing that there’s no actual false memory of the event, but rather the information just biases the participant’s answer.

EDIT: The strategic effects account could be a good theory to use to evaluate your methods. It can show that maybe this procedure isn’t great at showing the misinformation effect, so you could suggest modifications where there’s a more clear measurement of a false memory (like the broken glass modification).

If you’re aiming for an IB 5 or 6 for the IA, you can get by with just the schema theory or misinformation effect explanations. You could even get a 7, provided you do these well. However, the challenging explanation I’ve posed above is only for the brave and daring who are really aiming for a 7. And if you are I  recommend you read “A theoretical review of the misinformation effect” by Ayers and Reder (1998 Link) so you can get a full review of multiple theoretical explanations for the misinformation effect and choose the one that you like the best.

Explaining the relevance of the aim: If you are using the above explanation I’ve suggested which claims that there’s no false memory created, you have to be careful with how you explain the relevance of the aim. Most students link this to eye-witness testimony and the criminal justice system (e.g. it could be applied to improve the process of interviewing witnesses to increase accuracy of responses). This explanation still works. But you could go one step further and really impress your examiner by explaining that this is a relevant aim because it’s challenging the claims made in this now classic study by L&P. You can read how McClosky and Zarazoga (1985) applied this theory in their experiments here.

You can see more about basic common mistakes made in the IA and other IA videos with heaps of tips in our IB Psychology IA Playlist.

These are the common errors that I’ve come across and some ideas on how to avoid them. This is not an exhaustive list and if anyone has any other errors or tips for Loftus and Palmer, please write a comment. Or if you disagree with these points, please let me know.