Ethical guidelines for staff and students who plan to use the autismspectrum quotient (AQ, Baron-Cohen et al) in research projects.
The AQ is a 50-item questionnaire that asks participants to agree or disagree with a series of
statements that reflect the autism phenotype. As such it has been used to show that levels of
so-called “autistic-traits” vary in the general population, with some individuals showing high
levels of autistic traits, and other individuals showing lower levels.
The AQ is therefore a popular tool in developmental and individual differences research and
many papers have been published which show that individual differences in cognition,
perception and even neurological features are correlated with individual differences in
autistic traits. This document contains guidance for staff and students who plan to use the AQ
in their research, particularly with respect to particular ethical issues to be aware of.
Before you start, there are two important things to bear in mind.
1) The AQ can not (and should not) be used to diagnose autism. Even if someone has a
high score on this questionnaire, this does not mean that they have autism. Diagnosing
autism is a complex process that can only ever be carried out by experienced
clinicians. The AQ must never be used to identify whether someone has autism or not.
Results arising from the AQ must never be discussed with participants in the context
of whether or not they may have autism. The AQ is not an appropriate tool to support
such a conversation and must never be used as such.
2) If your research involves comparing people with high or low scores on the AQ, you
are not investigating “autism” per se. You are only investigating differences between
people with high or low scores on the AQ. It is more appropriate to frame your
research in terms of individual differences, where individual differences in traits and
behaviours that are associated with the autism phenotype, is the focus of interest.
The following guidelines have been prepared by the ethics committee to help you plan and
carry out your research project in-line with ethical principles.
1) Do not describe the questionnaire as measuring “autism”, or being used for diagnostic
purposes. As described above, the AQ does not measure autism, but rather individual
differences in some of the traits and preferences associated with the autistic
phenotype. As such, do not call the questionnaire that you administer to participants
the AQ. Instead, use the term “ a questionnaire which addresses personality traits and
preferences”.
2) Do not provide feedback about participants’ scores on the questionnaire. The
information sheet should indicate that experimenters are not able to provide individual
feedback from the research study.
3) Maintain confidentiality at all times. Ideally, participants’ responses will be
anonymised. If that is not possible then the experimenter must ensure complete
confidentiality.
4) It is possible that after completing the AQ, some participants may feel that they
identified particularly with a set of questions and recognise that the questionnaire is
focussed on behaviours associated with autism. Therefore, it is possible that after
completing the questionnaire a participant may ask the experiment for further
information / advice / support. Remember that your role here is as a researcher, and
you do not have the clinical background to comment on any individuals’ diagnosticstatus. In this situation it is appropriate to acknowledge that the AQ asks questions
which are associated with the autism phenotype but that it is not a diagnostic
instrument and therefore scores from the questionnaire can’t be used to infer anything
about diagnosis. Instead, you should direct participants to appropriate sources of
support. These include, the Autism Research Centre (Cambridge)
www.autismresearchcentre.com, and the National Autistic society www.nas.org for
more information. Any participant who feels that they require support should be
directed to the disability support service (DDSS) at the University of Sheffield. Alison
Jones provides specialist support for students with autism or Asperger’s. The
Disability Support main contact phone number is: 0114 2221303. The e-mail is:
disability.info@shef.ac.uk. If participants are not University of Sheffield students,
then they should be directed to talk to their GP if they are concerned. Their GP will
need to refer them to a clinical psychologist or psychiatrist for a formal assessment if
this is considered appropriate.
5) Given that no deception is involved, there is no need for a formal debrief after the use
of the AQ.
Examples of acceptable text that could be used to recruit participants are given below.
“We are looking for volunteers to complete a questionnaire which asks about preferences
and personality traits. The questionnaire should take approximately 5-10 minutes to
complete. The questionnaire is for research purposes only and assess patterns of traits in the
typical population. Please note that we will not be able to provide feedback about your scores
on the questionnaire.”
Some studies involve screening a large population first, and then inviting participants to take
part in further study on the basis of their scores (i.e. either high or low). This is acceptable,
providing that information is not presented in such a way that participants become concerned
about why they have been selected to participate. Example text to minimise distress in such
situations is:
“On the basis of your patterns of responses you may be invited to participate in further
research studies being conducted at the University of Sheffield Psychology department.
Participation in any future studies is entirely voluntary, and if we do invite you to participate
you will be under no obligation to do so.”
And, if participants are invited back on the basis of their scores…
“Many thanks for completing the online questionnaires about personal preferences and
personality traits. From your responses, we would like to invite you to participate in a
research study which investigates ____________________ (insert further details about the
particular study here).
You have been invited to participate based on the pattern of responses that you gave to these
questionnaires. Please note, these questionnaires have been coded automatically and you
were identified by the computer as providing a set of responses suitable for further study. We
are not able to provide feedback on your particular scores on the questionnaires.
There is no obligation to participate in this, or any other study. If you do not want to take
part there is no need to reply to this e-mail. If you do not want to hear from us again, please
reply with the word “remove” in the subject title. However, if you are interested in taking
part, please e-mail _________________ (insert name of researcher co-ordinating the study
here) for more information and to arrange a suitable time.”
If you are taking the approach of recruiting participants for further study on the basis of their
scores on the AQ then a good approach is to have one experimenter who identifies
participants who fit into the required range, and another experimenter to collect the data. The
first experimenter can identify the participants who meet inclusion criteria for the study, and
pass on contact details (but not AQ score) to the second experimenter. In this way, the second
experimenter does not know the participant’s AQ score while collecting data. This is an
ethical approach, and also a rigorous scientific approach, as it reduced the possibility that
experimenters may bias results if they are aware of the AQ score of the participant at the time
of collecting data. An example of a paper that has used this approach is Milne et al. 2013,
Neuropsychologia.
Elizabeth Milne, May 2016.