Raven’s Coloured Progressive Matrices: A useful tool to

match adults with Down syndrome and controls? Claudio Straccia and Koviljka Barisnikov

Child Clinical Neuropsychology Unit, Department of Psychology, University of Geneva, Switzerland

Corresponding author: Claudio Straccia, Child Clinical Neuropsychology Unit, University of Geneva, Uni Mail – 6164, CH-1211 Geneva 4, Switzerland, Phone +41 22 379 89 93, Fax +41 22 379 93 59 e-mail: Claudio.Straccia@unige.ch.

Background Studies on behavioural phenotypes often match participants on their cognitive abilities

Since persons with Down syndrome (DS) commonly present relatively preserved non-verbal reasoning

abilities associated to larger difficulties in verbal skills, general IQ is not an adequate matching variable

This cognitive dissociation is better controlled by matching participants on a non-verbal reasoning task and

by statistically controlling the influence of verbal skills (Straccia, et al., 2014)

The Raven’s Coloured Progressive Matrices (CPM, Raven et al., 1998) is often used to match participants with

DS to controls

While several studies suggest that children with DS present a specific error pattern (Gunn & Jarrold, 2004), other

studies find inconsistent results (Facon & Nuchadee, 2010)

Aim Since no studies have analyzed this

topic among adults with DS, we aim to

examine the usefulness of the CPM in

matching adults with and without DS

Participants: 48 adults with DS and 184 adults

with non-specific ID (NS) took part in the study

Material: the CPM is a non-verbal reasoning task

in which participant have to choose the piece

which correctly complete the presented pattern

(36 items, Cf. Figure 1)

Matching: the analyses were firstly conducted on

the two complete group and secondly by

individually matching participants with DS and NS

on the CPM raw score

Results: all participants (I) Results: matched groups (II)

0

10

20

30

40

50

60

70

Error A Error B Error C Error D

% o

f e

rro

rs

Error Analysis (I)

NS

SD

Figure 1. CPM item example

Figure 2. Items analysis formula

0

10

20

30

40

50

60

Error A Error B Error C Error D

% o

f e

rro

rs

Error Analysis (II)

NS

SD

Figure 3. Error analysis of the complete groups Figure 5. Error analysis of the matched groups

Figure 4. Items analysis of the complete groups Figure 6. Items analysis of the matched groups

Results (III) and Discussion

References

Error Analysis: the CPM provides four types of

error. We compared the two groups on the

proportions of each type of error

Items Analysis: this method consists in

computing a delta (Di) score for each item and a

principal axis (y = ax + b, Cf. Figure 2). A

difference greater than 1.5 delta units between

the principal axis and a delta point indicates a

significant differential functioning between the two

groups for a given item

Error A. Difference: The piece has no figure of any kind on it; the

figure shown is quiet irrelevant

Error B. Inadequate Individuation: The figure is contaminated by

irrelevancies or distortions; it combines figures irrelevantly; it is

whole or half the pattern to be completed

Error C. Repetition of the Pattern: Above and to the left,

immediately above or immediately to the left of the space to be

filled

Error D. Incomplete Correlate. The figure is wrongly oriented; it is

incomplete, but correct as far as it goes

Facon, B., & Nuchadee, M.-L. (2010). An item analysis of Raven's Colored Progressive Matrices among participants with Down syndrome. Research in Developmental Disabilities, 31(1), 243-249. doi: 10.1016/j.ridd.2009.09.011

Gunn, D. M., & Jarrold, C. (2004). Raven's matrices performance in Down syndrome: Evidence of unusual errors. Research in Developmental Disabilities, 25(5), 443-457. doi: 10.1016/j.ridd.2003.07.004

Raven, J. C., Court, J. H., & Raven, J. (1998). Progressive matrices - Colour. Oxford, UK: Psychologist Press.

Straccia, C., Baggio, S., & Barisnikov, K. (2014). Mental Illness, Behavior Problems, and Social Behavior in Adults With Down Syndrome. Journal of Mental Health Research in Intellectual Disabilities, 7(1), 74-90.

doi:10.1080/19315864.2012.741660

Comparing means: The adults with NS showed higher scores than the

adults with DS (t(98) = 3.545, p = .001)

Error Analysis: The two groups did not show any difference in the error

pattern, neither in matched nor in no matched conditions (Figures 3 and 5)

Items Analysis: The group with DS showed more difficulties to complete all

items, although this difference was not significant (< 1.5 delta units, Figure 4)

Items Analysis: In the matched conditions, the CPM items were equally

distributed close to the principal axis (Figure 6)

Inconsistent with Gunn & Jarrold (2004) findings, adults with DS presented

the same error pattern as their counter parts with NS

The Items Analysis results were consistent with previous studies (e.g. Facon

& Nuchadee, 2010), confirming that the items difficulty was equal in the two

groups and that it depended on global performance

Thus we can state that the CPM global score can be trustfully used to match

adults with DS and controls