Skip to main content
Top
Published in: Molecular Autism 1/2021

Open Access 01-12-2021 | Research

The Comprehensive Autistic Trait Inventory (CATI): development and validation of a new measure of autistic traits in the general population

Authors: Michael C. W. English, Gilles E. Gignac, Troy A. W. Visser, Andrew J. O. Whitehouse, James T. Enns, Murray T. Maybery

Published in: Molecular Autism | Issue 1/2021

Login to get access

Abstract

Background

Traits and characteristics qualitatively similar to those seen in diagnosed autism spectrum disorder can be found to varying degrees in the general population. To measure these traits and facilitate their use in autism research, several questionnaires have been developed that provide broad measures of autistic traits [e.g. Autism-Spectrum Quotient (AQ), Broad Autism Phenotype Questionnaire (BAPQ)]. However, since their development, our understanding of autism has grown considerably, and it is arguable that existing measures do not provide an ideal representation of the trait dimensions currently associated with autism. Our aim was to create a new measure of autistic traits that reflects our current understanding of autism, the Comprehensive Autism Trait Inventory (CATI).

Methods

In Study 1, 107 pilot items were administered to 1166 individuals in the general population and exploratory factor analysis of responses used to create the 42-item CATI comprising six subscales: Social Interactions, Communication, Social Camouflage, Repetitive Behaviours, Cognitive Rigidity, and Sensory Sensitivity. In Study 2, the CATI was administered to 1119 new individuals and confirmatory factor analysis used to verify the factor structure. The AQ and BAPQ were administered to validate the CATI, and additional autistic participants were recruited to compare the predictive ability of the measures. In Study 3, to validate the CATI subscales, the CATI was administered to 202 new individuals along with existing valid measures qualitatively similar to each CATI subscale.

Results

The CATI showed convergent validity at both the total-scale (r ≥ .79) and subscale level (r ≥ .68). The CATI also showed superior internal reliability for total-scale scores (α = .95) relative to the AQ (α = .90) and BAPQ (α = .94), consistently high reliability for subscales (α > .81), greater predictive ability for classifying autism (Youden’s Index = .62 vs .56–.59), and demonstrated measurement invariance for sex.

Limitations

Analyses of predictive ability for classifying autism depended upon self-reported diagnosis or identification of autism. The autistic sample was not large enough to test measurement invariance of autism diagnosis.

Conclusions

The CATI is a reliable and economical new measure that provides observations across a wide range of trait dimensions associated with autism, potentially precluding the need to administer multiple measures, and to our knowledge, the CATI is also the first broad measure of autistic traits to have dedicated subscales for social camouflage and sensory sensitivity.
Appendix
Available only for authorised users
Footnotes
1
Internal consistencies are also retrospectively reported for the social and non-social bifactors identified using confirmatory factor analysis in Study 2.
 
2
Where discretion was used to ignore certain items and select viable alternatives, the reasoning is outlined alongside the initial pattern matrix in Additional file 1: Table S2.
 
3
Readers may notice that the number of exclusions is considerably lower than what was reported for Study 1. We suspect that a certain portion of participants recruited through web-platforms such as Prolific Academic aim to sign up early to new studies and complete them as quickly as possible to earn monetary rewards. As Study 2 did not allow participants from Study 1 to participate, we hypothesise that most of the early, rapid responders were prevented from attempting the second study. We advise researchers to use attention checks and keep a record of the IDs of participants who fail these checks to inform their future studies on online platforms.
 
4
Similar patterns of group differences were found when t tests were conducted on the total-scale scores for the AQ, p = .89, d = .03, and the BAPQ, p = .29, d = .19.
 
5
For non-clinical samples, the scale’s authors recommend recoding any ‘4’ responses as a ‘3’.
 
Literature
1.
go back to reference Kanner L. Autistic disturbances of affective contact. Nerv Child. 1943;2:217–50. Kanner L. Autistic disturbances of affective contact. Nerv Child. 1943;2:217–50.
15.
go back to reference Ingersoll B, Wainer A. The broader autism phenotype. In: Volkmar FR, Rogers SJ, Paul R, Pelphrey KA, editors. Handbook of autism and pervasive developmental disorders. 4th ed. New Jersey: Wiley; 2014. p. 28–56. Ingersoll B, Wainer A. The broader autism phenotype. In: Volkmar FR, Rogers SJ, Paul R, Pelphrey KA, editors. Handbook of autism and pervasive developmental disorders. 4th ed. New Jersey: Wiley; 2014. p. 28–56.
19.
go back to reference American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 5th ed. Washington: American Psychiatric Publishing; 2013.CrossRef American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 5th ed. Washington: American Psychiatric Publishing; 2013.CrossRef
22.
go back to reference English MCW, Gignac GE, Visser TAW, Whitehouse AJO, Maybery MT. A comprehensive psychometric analysis of autism-spectrum quotient factor models using two large samples: model recommendations and the influence of divergent traits on total-scale scores. Autism Res. 2020;13(1):45–60. https://doi.org/10.1002/aur.2198.CrossRefPubMed English MCW, Gignac GE, Visser TAW, Whitehouse AJO, Maybery MT. A comprehensive psychometric analysis of autism-spectrum quotient factor models using two large samples: model recommendations and the influence of divergent traits on total-scale scores. Autism Res. 2020;13(1):45–60. https://​doi.​org/​10.​1002/​aur.​2198.CrossRefPubMed
23.
go back to reference Russell-Smith SN, Maybery MT, Bayliss DM. Relationships between autistic-like and schizotypy traits: an analysis using the Autism Spectrum Quotient and Oxford-Liverpool Inventory of Feelings and Experiences. Pers Individ Differ. 2011;51(2):128–32.CrossRef Russell-Smith SN, Maybery MT, Bayliss DM. Relationships between autistic-like and schizotypy traits: an analysis using the Autism Spectrum Quotient and Oxford-Liverpool Inventory of Feelings and Experiences. Pers Individ Differ. 2011;51(2):128–32.CrossRef
35.
go back to reference Nunnally JC, Bernstein IH. Psychometric theory (McGraw-Hill series in psychology). 3rd ed. New York: McGraw-Hill; 1994. Nunnally JC, Bernstein IH. Psychometric theory (McGraw-Hill series in psychology). 3rd ed. New York: McGraw-Hill; 1994.
42.
go back to reference Samejima F. Estimation of latent ability using a response pattern of graded scores. Psychom Monogr Suppl. 1969;34(4, Pt. 2):100. Samejima F. Estimation of latent ability using a response pattern of graded scores. Psychom Monogr Suppl. 1969;34(4, Pt. 2):100.
47.
go back to reference Gould J, Ashton-Smith J. Missed diagnosis or misdiagnosis? Girls and women on the autism spectrum. Good Autism Pract. 2011;12(1):34–41. Gould J, Ashton-Smith J. Missed diagnosis or misdiagnosis? Girls and women on the autism spectrum. Good Autism Pract. 2011;12(1):34–41.
Metadata
Title
The Comprehensive Autistic Trait Inventory (CATI): development and validation of a new measure of autistic traits in the general population
Authors
Michael C. W. English
Gilles E. Gignac
Troy A. W. Visser
Andrew J. O. Whitehouse
James T. Enns
Murray T. Maybery
Publication date
01-12-2021
Publisher
BioMed Central
Published in
Molecular Autism / Issue 1/2021
Electronic ISSN: 2040-2392
DOI
https://doi.org/10.1186/s13229-021-00445-7

Other articles of this Issue 1/2021

Molecular Autism 1/2021 Go to the issue