Method

3.2.1.1 Participants

One hundred and sixty three UK university students were recruited, aged between 18 and 50 years (M=21.32 years, SD=4.67; 5 female, 67 male, 1 unreported). Participants were undergraduate and postgraduate students recruited via the

university and social media. The first wave of participants (N=76) were recruited for my Master’s dissertation, and the remainder were recruited during my PhD. In order to increase the number of male participants, male non-Psychology students (N=20) were also specifically targeted via the university Notice Board. Psychology

undergraduates (N=120) received course credits in exchange for participation. Two participants scored at or above the clinical cut-off of 32 on the AQ and were removed from further analyses, resulting in a new sample (N=161) comprising 95 women and 65 men (1 unreported) with a mean age of 21.28 years (SD=4.69). The majority (N=136) were aged 18-22 years.

3.2.1.2 Materials

3.2.1.2.1 The RBQ-2A

The twenty items comprising the RBQ-2A are presented overleaf (Table 3.1) and these items are presented with full response options in Appendix 3 (pages 250- 253). For thirteen items, there are four available responses, corresponding to never or rarely (1), mild or occasional/one or more times daily (2), marked or notable/15 or

84

more times daily (3), and serious or severe/30 or more times daily (4), with additional information given regarding the effects of behaviour for eight items. For the

remaining seven items, only the first three response levels are offered.

Table 3-1 Full list of RBQ-2A items and number of responses per item.

Item Full question Shorthand Response

options 1 Do you like to arrange items in rows or patterns? Arrange 4 2 Do you repetitively fiddle with items? (e.g. spin, twiddle, bang, tap, twist,

or flick anything repeatedly?

Fiddle 4

3 Do you spin yourself around and around? Spin 4

4 Do you rock backwards and forwards, or side to side, either when sitting or when standing?

Rock 4

5 Do you pace or move around repetitively? (e.g. walk to and fro across a room, or around the same path in the garden?)

Pace 4

6 Do you make repetitive hand and/or finger movements? (e.g. flap, wave, or flick your hands or fingers repetitively?

Hand/ finger 4 7 Do you have a fascination with specific objects? (e.g. trains, road signs or

other things?)

Fascination 3 8 Do you like to look at objects from particular or unusual angles? Angles 3 9 Do you have a special interest in the smell of people or objects? Smell 3 10 Do you have a special interest in the feel of different surfaces? Feel 3 11 Do you have any special objects you like to carry around? Carry 3 12 Do you collect or hoard items of any sort? Collect 3 13 Do you insist on things at home remaining the same? (e.g. furniture

staying in the same place, things being kept in certain places, or arranged in certain ways?)

Home 4

14 Do you get upset about minor changes to objects? (e.g. flecks of dirt on your clothes, minor scratches on objects?)

Change 4

15 Do you insist that aspects of daily routine must remain the same? Routine 4 16 Do you insist on doing things in a certain way or re-doing things until they

are “just right”?

Redoing 4

17 Do you play the same music, game or video, or read the same book repeatedly?

TV/Music 4 18 Do you insist on wearing the same clothes or refuse to wear new clothes? Clothes 4 19 Do you insist on eating the same foods, or a very small range of foods, at

every meal?

Food 4

20 What sort of activity will you choose if you are left to occupy yourself? Activities 3

As mentioned in the previous chapter (page 67), item 20 is typically removed from factor analytic studies of the RBQ-2. This is because it has a non-quantitative response scale that is distinct from the other items (Leekam et al., 2007). Although all of the RBQ-2A’s response scales are ordinal, the first nineteen items focus on the frequency or severity of a behaviour and therefore have a natural order (e.g. one or more times daily is clearly lower than thirty or more times daily). In contrast, item 20 asks about the quality of activities (e.g. a restricted range of activities, or a flexible and

85

varied range of activities); whether or not these items represent the higher or lower end of a scale is determined by the researcher. That is, a high score could reflect more restricted behaviours, or a high score could reflect more flexible behaviours depending on the design of the survey. In the case of the RBQ-2A, high scores on this item reflect more restricted behaviours. As a result of its removal from factor analysis, item 20 is usually included only in the total score rather than the subscale scores. Therefore item 20 has been neglected in previous research despite containing a potentially rich source of information regarding RRBs; I shall return to this item in Chapter Four (page 119).

The most extreme responses are rarely endorsed and so in order to make all items comparable with each other, the two most extreme responses for items 1-6 and 13-19 are collapsed into one, resulting in a three-point scale in accordance with previous research (Leekam et al., 2007; Lidstone, Uljarević et al., 2014). Mean scores are used when analysing the RBQ-2A so as to counteract the effects of missing data, calculated by dividing the participant’s total score (out of 60) by the number of items actually answered, resulting in a mean total score between 1 and 3 for each

participant. Subscale scores may then be calculated according to previous research, or as is the case here by running PCA in order to determine components.

3.2.1.2.2 The AQ

The AQ (Baron-Cohen et al., 2001) is a self-report questionnaire assessing the presence of autistic features in the general population. It comprises 50 statements based on the original triad of impairments (social interaction, social communication and social imagination; Wing & Gould, 1979), and other aspects of cognitive processing in ASD and was originally divided into five subscales (imagination, social skills,

attention switching, attention to detail and communication) although this was not on the basis of empirical evidence. Each item is answered on a four-point scale ranging from definitely disagree to definitely agree. The questionnaire is then scored in a dichotomous manner, such that for half of the items the agree responses represent an autistic trait that is scored 1, and the disagree responses are scored 0; and vice versa for the remaining half of the items. The total number of 1s and 0s are then calculated and each participant receives a score out of 50, with higher scores indicating greater endorsement of autistic traits. In most studies, a score of 32 is considered the clinical

86

cut-off for ASD in accordance with the original paper (Baron-Cohen et al., 2001). Later research has recommended a more stringent cut-off of 26 (Woodbury-Smith,

Robinson, Wheelwright, & Baron-Cohen, 2005). Here I chose to implement the the original cut-off score of 32 in order to preserve a larger sample size and greater variation in responses.

In order to assess the correlations between the individual AQ subscales and the RBQ-2A, the full range of scores (from 1 to 4) were analysed rather than the traditional dichotomous scoring described above (e.g. Kloosterman et al., 2011; Stewart & Austin, 2009). This was to increase the amount of information captured by the AQ and so that the AQ responses were more comparable to the RBQ-2A. The necessary items were reversed so that a higher score on the AQ subscales indicated a higher level of autistic traits; that is, a higher score on the AQ imagination subscale indicates more difficulties with imagination.

3.2.1.3 Procedure, data screening and statistical analyses.

Ethical approval was obtained from the university’s School of Psychology Ethics Committee, and informed consent was obtained from the participants before they completed the questionnaires. The online questionnaires were presented on Google Documents, with the RBQ-2A presented first followed by the AQ10_{. The data were}

analysed using SPSS 20. The PCA was run according to the analytic strategy outlined in the previous chapter. The internal consistency for the whole scale and each of the resultant components was also assessed by calculating Cronbach’s alpha (α) values. The correlations between age and RBQ-2A score, and between AQ and RBQ-2A scores, were also assessed. Finally, the difference in scores between resultant sub-scales were also tested, along with the correlations between the sub-scales.