Files ending "ChoiceData" hold the data from the main experimental task, including the overall ICAR and AQ scores for each participants.
Files ending "AQScores" hold the responses for the short-form AQ.
Files ending "ICARscores" hold the ICAR scores for each participant.

These files hold the raw data, prior to exclusion based on attention-checks (participants who failed more than 2 catch trials were excluded -- see main text of paper). These data files only include participants who completed the task, whose ip address had not already been recorded in the data file, and who were eligible to participate (e.g., over 18 years of age).

Details regarding each file are given below.

AQChoiceData.csv:

Data for the study comparing high and low autism-quotient groups. Two versions were conducted; this file contains the combined data from both versions.

version: Identifies which version of the study the participant took part in. The different versions used different product categories but were otherwise identical. (See paper and supplemntary materials for details).
Pcp: unique participant identifier
Set: 0 to 25, unique identifier for the 26 choices the participant had to make (including 6 catch trials)
Trial: Trial number (counted from 0 = first trial)
Choice: Alternative chosen by the participant for that trial (A,B or D, D=Decoy)
TopAtt: the attribute that appeared on the top half of the display (see Figure 1 in paper)
BotAtt: the attribute that appeared on the lower half the display (see Figure 1 in paper)
Pos1,Pos2,and Pos3: Identify which option (choice alternative) occupies each of the 3 positions on the screen: 1=top, 2=left, 3=right (See Figure 1 in paper)
Target: identifies which of the alternatives (A or B) was the target (as defined by its relationship to the decoy). D was the target on catch trials.
Catch: identifies whether trial was a catch trial
Order: identifies the order in which the participant completed the experiment tasks: 1 = ICAR followed by Choice task; 2 = Choice task follwed by ICAR
Product: the product category for that trial
Adom: The attribute on which alternative A was dominant
Bdom: The attribute on which alternative B was dominant
Catchfail: the number of catch trials that the participant failed during the experiment
Dur: time in milliseconds to reach a decision (from appearance of choice display until"next" button clicked to register response; not a highly-accurate measure)
AQgroup: Whether participant was in the High or Low Autism Quotient group
ICAR: score on cognitive ability measure
AQScore: score on autism quotient
Age, Gender, Nationality: Demographic information. 
Aut: whether participant had a diagnosis of autism

AQChoiceData.csv:

Data for study comparing diagnosed autism with neurotypical controls

As above with the following additional columns/coding:

Order: 1 = AQ,ICAR,Choice; 2 = AQ,Choice,ICAR; 3 = Choice,ICAR,AQ; 4 = Choice,AQ,ICAR; 5 = ICAR,Choice,AQ; 6 = ICAR,AQ,Choice
ASC: Identifies whether participant in autism group or control group.
Comorbid: 1 = self-reported having a diagnosis of ADHD or OCD

AQStudyAQscores.csv and ASCStudyAQscores.csv:

Pcp: unique participant identifier
Item: unique question identifier (for details, see p.592 of Hoekstra, et al 2011)
Response: 1 = Definitely Agree, 2 = Slightly Agree, 3 = Slightly Disagree, 4 = Definitely disagree
Score: Range from 1 to 4. Scores match the response for non-autistic traits, and are reversed for autistic traits

The sum of the scores gives each particpant's autism quotient. Possible range: 28 to 112

AQStudyICARscores.csv and ASCStudyICARscores.csv

Pcp: unique participant identifier
Item: unique qustion identifier (for details, see Supplementary Materials for Condon & Revelle, 2014)
Response: participant's choice from available options (questions were multiple choice; response coding given in Condon & Revelle)
Icartype: one of four question types: VR= verbal reasoning; LN = Letter and number series; MX = Matrix reasoning; R3 = 3D Rotation
Answer: the correct answer
Correct: indicates whether response was correct

The sum of correct answers gives each participant's ICAR score. Possible range: 0 to 16

References

Condon, D. M., & Revelle, W. (2014). The international cognitive ability resource: Development and initial validation of a public-domain measure. Intelligence, 43, 52–64.
Hoekstra, R. A., Vinkhuyzen, A. A. E., Wheelwright, S., Bartels, M., Boomsma, D. I., Baron-Cohen, S., … van der Sluis, S. (2011). The construction and validation of an abridged version of the autism-spectrum quotient (AQ-Short). Journal of Autism and Developmental Disorders, 41(5), 589–596.