1

When less is more: poor discrimination but good colour

memory in Autism

Pamela Heaton1, Amanda Ludlow & D. Roberson2

1 Goldsmiths College, University of London, New Cross, London SE14 6NW, UK

2 University of Essex, Wivenhoe Park, Colchester CO4 3SQ,

Correspondence to:

Pamela Heaton, Ph.D.

Department of Psychology, Goldmsiths College,

University of London, New Cross, London SE14-6NW

UK.

Telephone: 020-7919-7884

Email: [email protected]

Abstract word count 94 words

Manuscript word count 3,930 words (excluding appendix and tables)

 2

Abstract: In two experiments children with autism and two groups of controls

matched for either chronological or non-verbal mental age were tested on tasks of

colour discrimination and memory. The results from experiment 1 showed

significantly poorer colour discrimination in children with autism in comparison to

typically developing chronological age matched controls. However, in experiment 2,

children with autism, retained unlabelled perceptual colour information to a

significantly higher level than either group of controls. The findings suggest that

enhanced performance on perceptual tasks relate to a reduced tendency to encode

verbal information in memory.

Introduction

Autism is a neurodevelopmental disorder of biological origin that is characterised by

deficits in socialisation, communication and imagination (American Psychiatric

Association, 1994). In addition to these diagnostic markers, approximately 70% of

diagnosed individuals also show corresponding global intellectual impairments

(Fombonne, 1999). However, for many of these individuals, specific, circumscribed

areas of cognitive functioning are preserved or even enhanced. For example,

unusually fine pitch discrimination has been noted in several experimental studies

(Bonnel, 2003; Heaton, Hermelin & Pring, 1998; Heaton, Pring & Hermelin, 1999;

Heaton, 2003; 2005) and this has lent support to models positing enhanced perceptual

functioning (Mottron & Burack, 2001: Mottron, Dawson, Soulières, Hubert &

Burack, 2006) or a bias to process and encode local information (Frith, 1989; Happé,

1999) in autism. The studies presented here will extend previous findings of enhanced

perceptual processing by testing perception and memory for colour in autism.

 3

An aspect of autism that is likely to have significant implications for perceptual and

cognitive information processing and organisation, but is largely unconsidered in

current theoretical accounts, relates to the delayed onset and wide variation in

language skills in diagnosed individuals (e.g. Kjelgaard & Tager-Flusberg, 2001).

Difficulties in generalisation and categorisation in autism have been attributed to

perceptual abnormalities, such as an increased tendency to process unique rather than

shared features in visual arrays (Plaisted, 2001). But recent studies of face processing

in autism (e.g. Deruelle, Ronda, Gepner & Tardif, 2004) show that autistic individuals

may use substantially different attentional and processing strategies from typically

developing children, at least for visual tasks. Moreover, recent investigations into

perceptual categorization (e.g. Roberson, Davidoff, Davies & Shapiro, 2004, 2005:

Pilling, Wiggett, Özgen & Davies, 2003; Goldstone, 1998) show that language plays a

central role in directing attention to those features of objects relevant for

categorization. Findings from the developmental literature show that young children’s

knowledge of object names results in preferential looking at both objects (Baldwin &

Markman, 1989) and pictures of objects (Schafer, Plunket & Harris, 1999). Further,

exhaustive sorting of objects into groups occurs within a week of entering the

‘naming spurt’ at around 18 months (Gopnik & Meltzoff, 1997).

Since Brown and Lenneberg (1954) first proposed a direct relationship between

naming and memory for colors, many studies have supported the proposal that

coloured stimuli are coded both perceptually and verbally by adults (Kay & Kempton,

1984; Lucy, 1992; Alvarado & Jameson, 2005; Özgen, 2004; Gilbert, Regier, Kay &
 4

Ivry, 2005; Pilling, Wiggett, Özgen & Davies, 2003; Roberson & Davidoff, 2000;

Witthoft et al., 2003). Studies of typically developing children’s colour naming and

memory have shown that discrimination and memory for colours is facilitated by

knowing the appropriate label for a colour, relative to colours they cannot name

(Sandhoffer & Smith, 2001; Kowalski & Zimiles, 2006; Braisby & Dockrell, 1999),

and this is true even when children know only one or two terms (Roberson, Davidoff,

Davies & Shapiro, 2004). This evidence suggests that language plays an important

role in shaping and modifying cognitive organisation in typical development.

However, language onset is significantly delayed in autism, and cognitive theories

propose that information processing is perceptually biased. The studies to be

described therefore explored the relative influences of verbal and perceptual

information on colour processing in autism. In experiment 1 we investigated colour

discrimination, and in experiment 2 we examined memory for colours in conditions

with and without verbal labels. As intelligence test score means for two of the

participants groups (Autism, MLD) were significantly below population norms, a pre-

test explored the extent to which each population could comprehend colour terms.

GENERAL METHOD

Participants

Thirteen children and adolescents with autism were recruited from a specialist school,

to which entry is dependent upon meeting criteria for autism outlined in DSM 1-V

(APA, 1994) or ICD-10 (World Health Organization, 1993). Testing using the Raven’s

Matrices task (Raven, Court, & Raven, 1992), a test of fluid intelligence, confirmed

that for the autism group, standardised scores were within the handicapped range.
 5

Measures of receptive vocabulary, using the British Picture Vocabulary Scales (BPVS;

Dunn, Whetton & Pintilie, 1997) were consistent with a group mean verbal mental

age (VMA) of 5 years. This profile, with non-verbal intelligence scores in the

moderate to mild range of handicap, and a large discrepancy between verbal mental

age (VMA) and chronological age (CA) is commonly associated with low-functioning

autism (LFA). There were two groups of control participants. The first were

intellectually unimpaired, were recruited from a mainstream UK state school, and

were individually matched to the participants with autism for chronological age

(Typically Developing group) (TD). The second group of controls were recruited from

a school for pupils with Moderate Learning Disabilities (MLD) and were matched to

the children with autism for chronological age and non-verbal IQ. Age and

psychometric data for the participant sample are shown in table 1.

(Table 1 about here)

Comparison of the psychometric data for the autism and two control groups revealed

no significant difference between chronological age measures (F (2,36)= .23). There

were significant between-group differences on Ravens Matrices scores (F (2,36) =

32.24, p<.001) and on verbal mental age measures (F (2,36) = 17.31, p<.001). Post-

hoc Bonferroni tests showed that TD controls had significantly higher Ravens

Matrices scores than MLD controls and participants with autism (both p<.05). Ravens

Matrices scores for MLD controls and participants with autism did not differ

significantly. TD participants had higher Verbal Mental Age scores than MLD
 6

controls and participants with autism (both p<.05), and MLD controls had higher

Verbal Mental Age scores than participants with autism (p<.05). Thus whilst

participants with autism were chronological age matched to both groups of controls,

and were also matched to MLD controls for non-verbal IQ, their language scores

(receptive vocabulary) were significantly poorer than those of both control groups.

Pre-test: Color Term Comprehension

Stimuli and Apparatus

11 colour tiles (prototypical examples of red, blue, yellow, orange, brown, pink,

purple, grey, black and white) were laid out on a table under natural daylight

conditions by a window. The experimenter read out the colour names, one at a time, in

random order and children were asked to point to the corresponding tile. Responses

were recorded and mean accuracy is shown in table 2.

(Table 2 about here)

A one-way Anova, carried out on the data failed to reveal a significant difference

between groups (F(2,36) = 2.38, p>.05).

Experiment 1: Testing perceptual discrimination.

Colours used in the following studies were glossy Munsell chips (Munsell, 1966).

Only the hue value of the colours varied, and chroma (colourfulness) and value

(lightness) were kept constant at Munsell levels 6/6. Good examples of the four
 7

primary hues (red, blue, green and yellow) at each of three hue levels were chosen

from the Munsell book and converted for the computer into L*u*v* co-ordinates

using Munsell conversion software. A full list of stimuli used in all three experiments

can be found in Appendix I. The selected coordinates were programmed into bit

images using CSS software and the stimuli were presented using e-prime software.

A portable spot chroma-meter was used to ensure that colours remained constant

across testing conditions.

Training trials: In eight trials (two for each of the four colours) participants saw

three colour patches, two of which (distractor patches) differed from the third (target

patches) in Munsell hue intervals of 8.5 or 9 steps (e.g. target = 1R6/6, distractors =

9.5R6/6 and 10R6/6). Thus the two distractors were always very similar to each other

and quite dissimilar from the target patch. Participants were instructed to point to the

most different patch (target) over as many trials as necessary to reach ceiling

performance. Verbal feedback was given throughout training trials.

Experimental trials: There were six trials of each colour (red, blue, green and

yellow) with the target stimuli differing from the two comparison stimuli by 1, 2 or

three Munsell steps. Each target appeared twice with each pair of distractors, making

a total of 24 trials, which were presented in random order.

Results.

The means and standard deviations for discrimination scores are shown in table 3.
 8

Place table 3 around here

A 2-tailed binomial test showed that the means for the (lowest scoring) autism group

was significantly above chance (p<.001). A one-way Anova showed a significant

between group effect (F(2,36) = 7.86, p<.001). Post-hoc Bonferroni tests showed that

TD control scores were significantly higher than scores for MLD controls and

participants with autism (both, p<.05). There was no significant difference between

scores for MLD control and autism groups.

Within group correlations were carried out on the discrimination scores, the

chronological and verbal mental age scores and the non-verbal IQ scores. Of these,

only the discrimination and verbal mental age correlations for the autism group (r = .

66, p<.05) and the TD controls (r = .72, p<.01) were significant.

Experiment 2: Testing colour memory

In the pre-test familiarisation phase of experiment 2, the same participants were

shown one of four animal pictures (taken from Snodgrass & Vanderwart, 1980),

paired with one of four different prototypical colour patches (red, green, blue or

yellow). The four animal-colour pairings (e.g. dog paired with red, cat paired with

blue, pig paired with green and rabbit paired with yellow) were shown four times

each, making a total of sixteen familiarisation trials. In the test phase the previously

exposed animals (dog, cat, pig and rabbit) were shown individually together with all

four of the colour patches (red, green, blue and yellow). Participants were instructed

to select, by pointing, the colour that had been associated with the animal in the
 9

familiarisation trials. Each participant had four attempts to match each animal/colour

(16 in all). Order of presentation of the trials was randomised. Responses were

recorded by the computer.

After a short break, the same familiarisation procedure was repeated for the second

phase of the experiment. However, in the test trials for phase 2, the pre-exposed

animals, were presented with three colour patches from the same colour category as

the patch paired with that animal in familiarisation trials (e.g. cat with three different

shades of blue). The position of the pre-exposed and distractor patches was

randomised across trials. A full list of stimuli is shown in Appendix I. Participants

were again asked to point to the colour patch that had been paired with the animal in

the familiarisation trial. Responses were recorded by computer.

Results

Table 4 shows the means and standard deviations from phase 1 and phase 2 of

experiment 2.

(Table 4 about here)

In phase 1 there were 4 response options, so the probability of chance performance

was 25%. In phase 2 there were 3 response options with a probability of chance

performance of 33%. To allow for differences in chance value scores, raw data were

converted into z scores. Z scores for the two phases of experiment 2 are shown in

table 5.
 10

(Table 5 about here)

As table 5 shows, all groups scored significantly above chance on phase 1. For phase

2 only autism group scores were significantly above chance.

Analysis of variance with Group (autism/MLDcontrols/TDcontrols) as the between

group variable and Condition (phase 1/phase 2) as the within group variable was

carried out on the data. This failed to reveal a significant main effect of Group (F

(2,36) = 2.71, p>.05). However, there was a significant main effect of Condition (F

(1,36) = 12.53, p<.001) and a significant group by condition interactions (F(2,36) =

8.56, p<.001) which is shown in figure 1 below.

Place figure 1 around here.

Correlates of verbal labelling: Although the participants with autism knew colour

terms (naming pre-test), and their colour discrimination (experiment 1) was not

significantly different to that of controls matched for age and non-verbal intelligence,

their performance on experiment 2 suggested that they relied less on verbal labels and

encoded more perceptual information. In order to determine which psychometric

variables are associated with information encoding strategies, memory scores (phase 1

& phase 2) were correlated with chronological age, verbal mental age, non-verbal IQ
 11

and discrimination ability (scores for experiment 1) for the whole subject sample (n=

39). Scores for phase 1 (where targets could be correctly identified by verbal coding

only) correlated with VMA (r=.58, p<.001) and discrimination ability (r = .54, p<.01).

For phase 2 scores (where successful performance relied on perceptual encoding)

there was a significant negative correlation with VMA (r = -.36, p<.05). No other

correlations reached significance. Within group correlations carried out on the control

group data (MLD & TD) all failed to reach significance, so the data were then pooled

and the correlations were repeated. These showed that for these participants without

autism, phase 1 scores correlated with VMA (r = .52, p<.01). No other correlations

were significant. Correlations carried out on the autism data were significant for

phase 1 and discrimination ability (experiment 1) (r = .62, p<.05). There was also a

significant negative correlation between phase 2 scores and VMA (r = -.55, p<.05).

These latter findings suggest that a bias to encode perceptual information in autism is

associated with low verbal mental age.

Discussion

In the pre-test phase of the experiments, participants were presented with a range of

eleven colours for identification. Whilst some individuals with autism and MLD made

some errors there was no significant difference between groups and the majority of

intellectually impaired participants named all colours correctly. TD children would be

expected to know all 11 basic colour terms by five years and the VMA means for all

groups were higher than this. The findings from experiment 1 were surprising given

that theoretical accounts of autism have proposed that enhanced perceptual

 12

discrimination is characteristic (WCC; EPF) and such abilities have been

demonstrated in other domains (e.g. pitch; Heaton, 2003; 2005; visuo-constructional;

Caron, Mottron, Berthiaume & Dawson; 2006). However, whilst participants with

autism performed at levels that were significantly better than chance, discrimination

scores were significantly poorer than those of CA matched typically developing

participants and very similar to participants matched for CA and non-verbal

intellectual impairment (MLD).

In experiment 2, animals and colour patches were presented for paired learning. In

phase 1 of the experiment, the colours with which the animals were associated were

both widely separated in perceptual space and were also from different name

categories. Therefore either attention to perceptual features or a naming strategy

should yield correct performance. In phase 2 of the experiment however, all three

choice alternatives received the same name, so a memory strategy based on naming

would not result in successful identification. Only by attending to the perceptual

qualities and ignoring the name classification of the presented stimuli would above-

chance performance be achieved. Children with autism performed above chance on

this condition (although controls did not) indicating that they remembered the exact

shades of the paired colours, rather than relying on the category name. Whilst all of

these participants knew the relevant colour names, these appeared to influence their

colour perception to a far lesser degree than that of children with similar levels of

non-verbal intelligence, but without autism.

 13

The correlations carried out on the data from the experiments revealed interesting

differences in patterns of performance between groups and across the different

conditions. Within the typically developing group, the most verbally able children

showed the highest levels of discrimination on experiment 1 and phase 1 of

experiment 2. Whilst discrimination in experiment 1 did not correlate with VMA for

the MLD controls, the phase 1 scores and VMA correlation approached significance.

Whilst MLD scores were poorer than those of TD participants, they resembled them

in showing a primary reliance on verbal labels in memory and chance level

performance when recognition depended on encoded perceptual information.

The low levels of performance, observed in participants without autism in phase 2 of

experiment 2, are surprising given empirical findings showing that colour stimuli are

encoded both perceptually and verbally by adults (e.g. Robeson & Davidoff, 2000;

Pilling et al., 2003). However, in experiment 2, attention was directed to both colours

and animals, and this may have reduced perceptual encoding of colour information.

Further, familarisation trials and test trials for phase 1 may have served to reinforce

the use of a verbal labelling strategy, which resulted in chance performance on phase

two.

The correlations carried out on the autism data showed that like TD participants,

discrimination performance on experiment 1 correlated with VMA. This finding is

interesting in that much of the empirical work that has supported enhanced perception

theories has been carried out with intellectually able individuals with autism.

However, in experiment 1, the upper range for discrimination scores was still lower
 14

than for the MLD group (19 MLD; 17 LFA). As was the case for the participants

without autism, there was a positive and significant correlation between VMA and

scores for phase 1 of experiment 2. However for phase 2, where performance was not

at chance, performance scores and VMA correlations were significant and negative,

showing that children with autism and low VMA retained unusually high levels of

perceptual information. This finding is important in providing evidence for a direct

link between perception and deficits in language in autism.

At the early, pre-linguistic stages of typical development, information processing

appears to be characterised by attention to perceptual features. For example, French,

Mareschal, Mermillod & Quinn (2004) were able to manipulate categorisation

performance in 3- 4 month old infants by making small changes to specific visual

features. However, whilst these changes influenced category re-allocation (e.g. cat to

dog category) in infants, such effects were not seen in adults (Rakison, 2000). The

probability that such a developmental shift in attentional processing is linked to the

onset of language is supported by evidence showing increased attention to objects and

pictures of objects for which verbal labels are available (Baldwin & Markham, 1989;

Schafer et al., 1999), and the close link between the onset of spontaneous exhaustive

sorting of object into groups and the naming spurt at 18 months (Gopnik & Meltzoff,

1997). If, as this evidence suggests, an early interplay between perceptual and

linguistic processes is an important characteristic of typical developmental trajectories

for the development of concepts and categories, then this raises important questions

about qualitatively atypical information processing in children for whom language has

been markedly delayed. Whilst most of the autistic participants knew colour names,
 15

these influenced information processing to a much smaller degree than children

without autism. For children with autism, the increased salience of perceptual

information reflects the impoverished role that language plays in cognitive

development. Late language competence may force autistic individuals to adopt novel

strategies in a wide range of perceptual categorisation tasks.

References

Alvarado, N. & Jameson,K.A. (2005). Confidence Judgments on Color Category Best

Exemplars, Cross-Cultural Research. 39, 134-158

American Psychiatric Association. (1994). Diagnostic and statistical manual of

mental disorders, 4th ed (DSM-IV). Washington, DC: APA.

Baldwin, D.A. & Markman, E.M. (1989) Establishing word-object relations: A

first step. Child Development, 60, 381-398.

Belmonte, M.K., Allen, G., Beckel-Mitchener, A., Boulanger, L.M., Carper, R.A. &

Webb, S.J. (2004). Autism and abnormal development of brain Connectivity, The

Journal of Neuroscience, 24 (42), 9228-9231;

Bonnel, A.C., Mottron, L., Peretz, I., Trudel, M., Gallun, E., & Bonnel, A.M. (2003).

Enhanced pitch sensitivity in individuals with autism: A signal detection analysis.

 16

Journal of Cognitive Neuroscience, 15, 226-235.

Bornstein, M.H. & Korda, N.O. (1984) Discrimination and matching within and

between hues measured by reaction times: Some implications for categorical

perception and levels of information processing. Psychological Research, 46,

207-222.

Braisby, N.,& Dockrell, J. (1999). Why is color naming difficult? Journal of Child

Language, 26, 23-47.

Brown, R., & Lenneberg, E. (1954). A study in language and cognition. Journal of

Abnormal and Social Psychology, 49, 454-462.

Caron, M-J., Mottron, L., Berthiaume, C. & Dawson, M. (2006). Cognitive

mechanisms, specificity and neural underpinnings of visuospatial peaks in autism.

Brain, 129, 1789 – 1802.

Deruelle, C., Rondan, C., Gepner, B., & Tardif, C. (2004). Spatial Frequency and Face

Processing in Children with Autism and Asperger Syndrome. Journal of Autism and

Developmental Disorders, 34, (2), 199-210.

Dunn, L.M., Whetton, C., & Pintilie, D. (1997). British picture vocabulary scale.

Windsor: NFER-Nelson.
 17

Fombonne, E. (1999). The epidemiology of autism: a review. Psychological Medicine,

29, 769-786.

French, R. M., Mareschal, D., Mermillod, M. & Quinn, P. (2004) The role of bottom-

up processing in perceptual categorization by 3- to 4-month old infants: Simulations

and data. Journal of Experimental Psychology: General, 133, 382-397.

Frith, U. (1989). Autism: Explaining the enigma. Oxford: Blackwell.

Aubrey Gilbert, Terry Regier, Paul Kay, and Richard Ivry (2006). Whorf hypothesis is

supported in the right visual field but not the left. Proceedings of the National

Academy of Sciences 103(2), 489-494.

Goldstone, R. L. (1998) Perceptual Learning. Annual Review of Psychology, 49,

585-612.

Gopnik, A., & Melzoff, A.N. (1997). Words, thoughts and theories. Cambridge, MA:

MIT press.

Happé, F. (1999). Autism: Cognitive deficit or cognitive style? Trends in Cognitive

Sciences, 3, 216-222.

Heaton, P. (2003). Pitch memory, labeling and disembedding in autism. Journal of

Child Psychology and Psychiatry, 44, (4), 543 – 551.

 18

Heaton, P. (2005) Interval and contour processing in autism. Journal of Autism and

Developmental Disorders, 35, (6), 787 – 793.

Heaton, P., Hermelin, B. & Pring, L. (1998). Autism and pitch processing: A precursor

for savant musical ability? Music Perception, 154, 291 - 305.

Heaton, P., Pring, L. & Hermelin, B. (1999). A Pseudo-savant: A case of exceptional

musical splinter skills. Neurocase, 5, 503 – 509.

Kay, P.& Kempton, W. (1984). What is the Sapir-Whorf hypothesis? American

Anthropologist, 86, 65-78.

Kjelgaard, M., & Tager-Flusberg, H. (2001). An inverstigation of language

impairment in autism: Implications for genetic subgroups. Language and Cognitive

Processes, 16, 287-308.

Kowalski, K. & Zimiles, H. (2006). The relation between children’s conceptual

functioning with colour and colour term acquisition. Journal of Experimental Child

Psychology, 94, (4), 301-21.

 19

Lucy, J. A. (1992). Language and diversity of thought: A reformulation of the

linguistic relativity hypothesis. Cambridge, UK: Cambridge University Press.

Mottron, L., & Burack, J. (2001). Enhanced perceptual functioning in the

development of persons with autism. In J.A. Burack, T. Charman, N. Yirmiya, & P.R.

Zelazo (Eds.), The development of autism: Perspectives from theory and research (pp.

131-148). Mahwah, NJ: Erlbaum.

Mottron, L., Dawson, M., Soulières, I.,Hubert, B. & Burack, J. (2006). Enhanced

Perceptual Functioning in Autism: An update and eight principles of autistic

perception. Journal of Autism and Developmental Disorders, 36, (1), 27 – 43.

Munsell (1966). Munsell book of color. Munsell Color Company. Baltimore.

Özgen, E. (2004). Language, Learning, and Color Perception

Current Directions in Psychological Science, 13, (3), 95–98.

Özgen, E., & Davies, I. R. L. (2002). Acquisition of categorical colour perception: A

perceptual learning approach to the linguistic relativity hypothesis. Journal of

Experimental Psychology General, 131, 477-493.

Pilling, M., Wiggert, A., Özgen, E., & Davies, I. R. L. (2003). Is colour ‘categorical

perception’ really perceptual? Memory and Cognition, 31, 538-551.

 20

Plaisted, K. (2001). Reduced generalization in autism: An alternative to weak central

coherence. In J. Burack, T. Charman, N. Yirmiya & P.R. Zelazo (Eds.). The

Development of Autism: Perspectives from Theory and Research (pp. 149-169).

Mahwah, NJ: Erlbaum.

Rakison, D.H. (2000). When a Rose Is Just a Rose: The Illusion of Taxonomies in

Infant Categorization. Infancy, 1, (1), 77 – 90.

Raven, J.C., Court, J.H., & Raven, J. (1992). Standard progressive matrices. OUP:

Oxford Psychologists Press.

Roberson, D., & Davidoff, J. (2000). The ‘Categorical Perception’ of colours and

facial expressions: the effect of verbal interference. Memory and Cognition, 28,

977-986.

Roberson. D., Davies, I. R. L., & Davidoff, J. (2000). Colour categories are not

universals: Replications and new evidence from a stone-age culture. Journal of

Experimental Psychology General, 29, (3), 369-398.

Roberson, R., Davidoff, J., Davies, I., & Shapiro, L. (2004). The development of

colour categories in two languages: a longitudinal study. Journal of Experimental

Psychology: General, 133, 554-571.

Roberson, D., Davidoff, J., Davies, I.R.L. & Shapiro, L. (2005) Colour categories in
 21

Himba: Evidence for the cultural relativity hypothesis. Cognitive Psychology, 50, 378-

411.

Sandhofer, C. M., & Smith, L. B. (2001). Why children learn color and size words so

differently: Evidence from adults’ learning of artificial terms. Journal of Experimental

Psychology: General, 130, 600-620.

Schafer G., Plunkett K. & Harris P.L. (1999). What’s in a name? Lexical knowledge

drives infants’ visual preferences in the absence of referential input, Developmental

Science, 2, (2), 187-194.

Snodgrass, G., and Vanderwart, M. (1980). A standardised set of 260 pictures: norms

for name agreement, image agreement, familiarity and visual complexity. Journal of

Experimental Psychology: Human Learning and Memory, 6, 174-215.

Witthoft, N, Winawer, J, Wu, L, Frank, M, Wade, A, Boroditsky, L. (2003). Effects of

language on color discriminability. Proceedings of the 25th Annual Meeting of the

Cognitive Science Society.

World Health Organisation (1993). The ICD-10 Classification of Mental and

Behavioural Disorders. Geneva: World Health Organisation

 22

Acknowledgements: This work was supported by EU grant (12984) Stages in the

Evolution and Development of Sign Use (SEDSU). We would also like to thank all

our participants and their teachers and parents.

Appendix 1
Stimuli used in Experiment 1

Discrimination Task
Practice Trials

1R6/6, 9.5R6/6, 10R6/6

1B6/6, 9.5B6/6, 10B6/6
1G6/6, 9.5G6/6, 10G6/6
1Y6/6, 9.5Y6/6, 10Y6/6

Test Trials
2.5B6/6, 5B6/6, 7.5B6/6, 10B6/6
2.5Y6/6, 5Y6/6, 7.5Y6/6, 10Y6/6
2.5R6/6, 5R6/6, 7.5R6/6, 10R6/6
2.5G6/6, 5G6/6, 7.5G6/6, 10G6/6

Memory Task
Part One

5R6/6, 5B6/6, 5G6/6. 5Y6/6

Part Two
5R6/6, 1R6/6, 9R6/6
5B6/6, 1B6/6, 9B6/6
5G6/6, 1G6/6, 9G6/6
5Y6/6, 1Y6/6, 9G6/6
 23

Table 1: Psychometric data for participants with autism and their controls.

Group Chron. Age Ravens Matrices Verb. Ment. Age

Mean sd Mean sd Mean sd
Autism (n=13) 11.4 2.46 62.15 6.82 5.38 .97
MLD Controls (n=13) 11.5 2.07 65.54 7.73 7.06 .97
TD Controls (n=13) 11.0 2.44 89.15 14.42 8.76 3.11

Table 2. Mean number of colours correctly identified in response to the

appropriate colour term at pre-test.

Group Mean s.d.

(max=11)

Autism 10.23 1.48

MLD Controls 10.77 .59

TD Controls 11.00 .0

Table 3. Mean number of correct colour discriminations in experiment 2.

Group Total correct discrimination

scores (max = 24)
Mean s.d.
Autism .53 3.23
MLD Controls 13.84 4.41
TD Controls 18.23 3.76

Table 4: Mean memory accuracy and standard deviations for the participants
with Autism and their controls on phase 1 and phase 2 of experiment 2.
 24

Group Phase 1 Phase 2

(max = 16) (max = 16)
mean sd mean sd
Autism 5.31 3.2 6.62 1.85

MLD Controls 6.92 4.59 3.92 1.44

TD Controls 10.69 5.2 4.54 2.36

Table 5: Z scores for phase 1 and 2 of experiment 2.

Group Phase 1 Phase 2

Z scores Z scores

LFA .92 2.47

MLD Controls .62 -1.87

TD Controls 13.9 -1.52

Figure 1. Interaction in experiment 2.

25
 26

All in-text references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.