It is important to remember that your goal is to test a theory - not simply to replicate an experiment. The theory should be explained in some detail.  Having a summary of the original study is not required - and in fact, this will help you with your word count. You should absolutely not include two or three studies relevant to the theory in the introduction; this was a requirement of the former curriculum and is no longer assessed - it just eats up your word count....

A well-written aim has the following characteristics:

• The independent and dependent can be recognized, even though they may not be operationalized.
• The target population to be studied is clearly identified.

Simply saying that your aim is to "replicate someone's study" will guarantee you low marks for this assessment criteria.

When explaining your study, you should be careful not to make simple statements like "it saved us time."  The IB gives you 30 hours to carry out the experiment; it is unlikely that your experiment is going to take up a lot of time either way. 

A common error is to say that you used a "random opportunity sample."  There is no such thing.  If it is an opportunity sample, it cannot be random. You could say that participants were "randomly selected from an opportunity sample" or that they were "randomly allocated to conditions." But it needs to be clear which is the case.

Trying to get your participants to "behave" is not a control.  Controls should be of variables that could influence the outcome of the study - but cheating is not one of them.

The examples above have many common errors.  These include:

• If using medians, you cannot use standard deviation.  You would have to use the IQR (interquartile range)
• If using a Likert scale, you have ordinal data.  You cannot take a mean as the level of data is too low.  The median is the most appropriate.
• T-tests should not be applied in samples less than 30.  If you still want to do a t-test, you need to carry out a test of homogeneity.
• When the data is nominal - yes/no - then the test is a chi squared. 

Please note, the IB strongly discourages students from carrying out a test with more than one level of the IV.  If you insist on doing this, please be aware that you must word your hypotheses correctly and carry out post-hoc testing.  If you don't know what that means, it is a good idea to keep it simple.

Once again, the focus is on the theory - not on the original study.  If your results do not support the theory, you may want to discuss how your study differs from the original and why this may have had an effect on the findings - but be careful to still link this to the theory.  For example, unlike in the original Loftus and Palmer study, our sample was made up of students who did not have drivers' licenses.  This could mean that they didn't have the same experience - and thus, schema - as the participants in the original study.

It is important to remember that ethics are not a strength of your study.  They are requirement.

Even though the appendices are not directly assessed, it is important that you include all relevant material.

• No raw data means that there is no proof you actually did the experiment. This is flagged as potential academic dishonesty.  The minimum penalty is four marks off the analysis.
• Failure to include a full print-out or calculations of the inferential statistics results in a penalty of 3 marks for the analysis.
• If you have your graph in the appendices but not in the body of the report - it is not assessed.
• Failure to include the blank consent form and debriefing notes means that there is no evidence that you met ethical standards.  This could result in your not receiving credit for your IA.
• Your procedure must be in the body of the report.  The procedure is used to assess the exploration, the analysis and the evaluation.  If you say that you did an independent samples design but you have not included the procedure, there is no way to know if this is actually correct - and your exploration, analysis and evaluation may all be penalized.

All of the points above are problematic - but not manipulating the IV is not an ethical problem.  If you do not manipulate the IV then the study is not experimental and may not earn credit.

It is important that you do not include any information that identifies you, your school or your participants.  In addition, you must include evidence that you worked as part of a group.  IAs without group numbers (or, in the case of non-IB students, some other form of identification), may result in no credit for the IA.