SAMPLE CHARACTERISTICS

3.1 A total of 88 people participated in the study, 49 appearing before Bourke Local Court and 38 before Brewarrina (for one subject this piece of data was missing). Full data sets for all sections of the survey instruments were not available for all participants, for a number of reasons:

(a) participants were sometimes interrupted part of the way through the procedure, either for a conference with their lawyer, or to be summoned into the court room. The researchers were skilled at following up these subjects, but some missing data were inevitable; and

(b) some subjects chose not to answer particular questions.

3.2 Throughout the report, the people who participated in the study are referred to as participants, respondents or subjects. Results will be compared with the results obtained in the previous court cohort study (RR 4).¹

DEMOGRAPHIC CHARACTERISTICS OF THE SAMPLE

(a) Gender

3.3 The sample consisted of 72 males (81.8 per cent) and 15 females (17.0 per cent) with one person not responding to the question. New South Wales local court statistics for 1994 indicate that 15.4 per cent of persons found guilty in matters which were finalised were women.² Thus, the gender distribution is probably consistent with New South Wales local courts generally.

(b) Ethnicity

3.4 The majority of respondents were Aboriginal (N=65; 73.9 per cent of those answering the question), and one was Islander (1.1 per cent). Fourteen respondents (15.9 per cent) indicated that they belonged to neither of these groups, and eight people did not respond.

3.5 Most respondents had been born in Australia (N=77; 92.8 per cent of those answering the question) and 6.0 per cent (N=5) indicated that they had been born elsewhere. This figure is lower than the 20 per cent of Australians, and of the New South Wales population who were born overseas.³ Only two subjects completed the item requesting information about where they had been born outside Australia, one indicating place of birth as New Zealand, and the other as Poland.

(c) Education

3.6 In this sample, 8 per cent had completed Year 12 at high school, 4 per cent fewer than in RR 4. A cumulative percentage of 84.7 per cent had completed Year 10, as shown in Table 1. This result is very similar to the 81 per cent who reported that they had completed Year 10 in RR 4, and is probably related to the legal age at which young people are permitted to leave school.

Table 1: Highest grade completed

Grade	Frequency	Cumulative percentage
2	2	2.4
3	5	8.2
4	7	16.5
5	2	18.8
6	5	24.7
7	2	27.1
8	10	38.8
9	26	69.4
10	13	84.7
11	6	91.8
12	7	100.0
-	N = 85	-

3.7 It is notable that 24.7 per cent of this sample had completed school only up until Grade 6, compared with 3 per cent of the sample in RR 4. This finding could reflect the difficulties faced by some children attending high school in rural areas, including distance of school from home, frequent family moves, or financial constraints preventing outlying students from being able to board in a country town in order to be able to attend high school.

3.8 The schooling history of the sample is shown in Table 2.

Table 2: Number of schools attended

No of schools	Frequency	Cumulative percentage
1	19	26.0
2	26	61.6
3	15	82.2
4	3	86.3
5 or more	10	100
	N = 73

3.9 Of the 19 subjects who reported attending only one school, only four left school at or before Grade 6, whereas the remainder left school in Years 8, 9 or 10, dispelling the idea that attending only one school is associated with leaving school after primary school. Other reasons for attending only one school could include:

(a) attendance at a private school which offers both junior and senior levels - possibly fairly rare in this sample;

(b) attendance at a special school for students with disabilities or problems, which was ungraded in terms of primary and secondary differentiation;

(c) attendance at only one school, with the student undertaking the remainder of schooling through correspondence, or informal education; and

(d) attendance at only one school because of receiving schooling during time spent in a juvenile institution.

3.10 Students who attended two schools and who did not fall into category (a) above, might have had the “ideal” educational history, that is, one primary school and one secondary school. Only 35.6 per cent of the sample had experienced this possibly ideal situation. A further 13.7 per cent reported the other end of the spectrum of education, having attended five or more schools. The greatest number of different schools reported was ten, by one respondent. The figures at the high end of the scale are lower than those reported in RR 4, where 20 per cent indicated that they had attended five or more schools, the largest number given being 20 schools.

3.11 Those reporting that they had attended a special class or school formed 12.5 per cent of the sample (N=11), with 5.7 per cent specifying that they had been in an OA or OF class (specialist classes for students with an intellectual disability in the New South Wales educational system, although now other labels are attached to special classes).

(d) Employment

3.12 Unemployed people were vastly over-represented in the sample compared with the general population; 62.5 per cent (N=55) stated that they were unemployed, a figure which was slightly higher than that found in RR 4 (58 per cent). In May 1994, 9.5 per cent of the New South Wales population were unemployed (using the Australian Bureau of Statistics definition of unemployment).⁴ Teenagers recorded the highest unemployment rate at 21.4 per cent with those aged 20-24 years recording the second highest rate at 13.5 per cent. In 1991, Bourke Shire had an unemployment rate of 10.2 per cent, slightly down from a high of 12.2 per cent in 1987; in Brewarrina Shire the 1991 unemployment rate was 10.5 per cent, down from a high of 23.9 per cent in 1987.⁵ The Australian Bureau of Statistics now includes the Northern, Far West, North West and Central West areas of New South Wales as one region, the unemployment rate for that region in May 1994 being 6.8 per cent, and for the May 1995 quarter being 7.6 per cent.⁶ The high rate of unemployment in both court cohort studies indicates that for the section of the community appearing before courts, unemployment is clearly a significant factor.

3.13 Fewer men (33.3 per cent) than women (42.9 per cent) were employed, although the difference is not statistically significant. Of those who were unemployed, 85.2 per cent were men, and 14.8 per cent were women, reflecting the M:F ratio of the sample; these figures were not significantly different from RR 4.

3.14 The categories of employment nominated by more than one subject were Commonwealth Development Employment Programme (N=6), labourer or road worker (N=3) and shire worker (N=2). Other occupations mentioned included shearer, counsellor, social worker, boner, gardener, postal worker, nurse, garbage worker, HIV/STD education officer, receptionist, police, shop assistant, and project supervisor. Very few of the respondents indicated that they belonged to professional or semi-professional categories of employment.

3.15 Seventeen per cent (N=15) of the total sample indicated that they were working full-time, and 23.9 per cent (N=21) indicated they were working part-time. The figure for full time employment as a proportion of those who were employed showed that 41.7 per cent were fulltime. In RR 4, 80 per cent of those working were in fulltime employment. The discrepancy between the two samples may reflect the difficulty of obtaining full time employment in a depressed and drought stricken rural sector. In New South Wales, 23.1 per cent of employed persons work part time.⁷

3.16 The total number of jobs is greater than the proportion indicating that they were employed, probably because some people had two jobs. Not all gave the numbers of hours worked per week, but for those who did the results are presented below in Table 3.

Table 3: Hours of work

Hours of work per week	Number	Percentage
Less than 10	1	3.4
10-19	6	20.6
20-29	6	20.6
30-39	4	13.8
40-49	9	30.9
50+	3	10.2
	N = 29	99.5

(e) Social Security

3.17 A total of 61.4 per cent (N=54) of the sample indicated that they were in receipt of a pension or benefit, a figure which is not significantly different from RR 4 (57 per cent). Table 4 shows the proportions in receipt of different types of benefit (one subject in receipt of a benefit did not specify the type).

Table 4: Pension or Benefit

Type of benefit	N	Percentage
Unemployment	14	26.4
Job Search	21	39.6
Disability Support	7	13.2
Sole/supporting parent	8	15.1
Other (Age, carer)	3	5.7
	N = 53	100.0

3.18 In comparison with RR 4, the proportion of the sample receiving unemployment or job search allowances is similar, whereas the number of those receiving disability support pensions or supporting parents benefits is greater.

3.19 A total of 17 respondents indicated that they had received a pension or benefit in the past; thus 79.5 per cent had at one time received a pension or benefit, predominantly unemployment benefits. Only six of the respondents in the borderline or intellectually disabled categories were in receipt of a Disability Support Pension, whereas 15 received unemployment or Job Search allowances, one a sole parent allowance, and one an aged pension. Thus, type of pension is not an accurate identifying characteristic for people with an intellectual disability.

(f) Age

3.20 The following Table 5 shows the age range of the sample compared with the age distribution of persons appearing before New South Wales local courts in 1994.

Table 5: Age range

Age	Frequency	Percentage	1994 stats %*
18 or less	5	5.7	6.2
19	2	2.3	6.2
20-24	25	28.4	26.9
25-29	19	21.6	17.6
30-39	21	23.8	24.0
40-49	9	10.1	11.1
50 plus	7	7.9	7.9
	88	99.8	N = 84816

(*See NSW Bureau of Crime Statistics and Research New South Wales Criminal Court Statistics 1994, Table 1.12.)

3.21 In relation to the New South Wales local court population, it can be seen that this sample is slightly under-represented in the under 19 years age range, and slightly over-represented in the 20-29 years range, but the differences are not statistically significant.

3.22 The age distribution of the sample may contribute towards the unemployment figures, as 36.4 per cent are aged 24 years or less, and as was mentioned above, unemployment is highest in these lower age categories.

3.23 The mean age of the sample is 30.7 years (standard deviation 11.0 years).

OFFENCES

3.24 Most of the responses to the question concerning reasons for being at court indicated that attendance at court was because of having been charged with an offence (88.5 per cent, N=77) or having breached an order; 29.6 per cent (N= 24) were charged with more than one offence (mostly two or three offences, but the highest number recorded was 12).

3.25 Table 6 shows the most serious charge reported by participants, categorised according to standard offence code categories.⁸

Table 6: Offence category of most serious charge

Category	Frequency	Percentage	1994 stats %*
Offences against the person	31	46.9	14.2
Break & enter, burglary etc	8	21.1	18.3
Property damage	-	-	4.5
Offences against good order	15	22.6	18.3
Drug offences	-	-	10.1
Driving, motor vehicle	10	15.2	33.9
Other	2	3.0	0.6
	66	99.8	N = 135262

(*See NSW Bureau of Crime Statistics and Research New South Wales Criminal Court Statistics 1994, Table 1.1, number of persons charged.)

3.26 The most frequent single charge was “other assault” which includes assault, aggravated assault, assault with weapons, assault with intent to resist arrest and assault in company (39.4 per cent of those responding to this question). This was the second most frequent charge in New South Wales local courts in 1994 (11 per cent of persons charged), the most frequent charge being “other driving” offences.⁹ In comparison with RR 4, this study group had a far higher proportion with respect to offences against the person (46.5 per cent, compared with 25.3 per cent). Unlike RR 4, however, this group were under-represented for drug offences. Both samples were under-represented on driving and motor vehicle offences.

3.27 Amongst ancillary charges nominated by respondents, offences against the person again formed the most frequent category (22.7 per cent) followed by offences against good order (18.1 per cent). Other research has found that people with an intellectual disability are most frequently convicted of offences against the person.¹⁰

3.28 Seventy participants (84.3 per cent of those responding to the question, and 79.5 per cent of the total sample) indicated a history of previous offences. Not all participants answered the question pertaining to the number of previous offences. Nevertheless, Table 7 shows the number of previous offences stated by 50 respondents.

Table 7: Number of previous offences

Number	Frequency	Percentage	Cumulative %	Previous study %
1	5	10.0	10.0	24
2	8	16.0	26.0	19
3	3	6.0	32.0	4
4	2	4.0	36.0	9
5	3	6.0	42.0	9
6	4	8.0	50.0	3
More than 6	25	50.0	100.0	33
	N = 50			N = 70

3.29 These results indicate that in this sample, respondents were more likely to have had six or more previous offences when compared with RR 4. Past offences were predominantly offences against the person, or against good order.

LEGAL REPRESENTATION

3.30 A total of 12.0 per cent of participants indicated that they did not have a lawyer; 81.9 per cent had Legal Aid representation; and 6.0 per cent had a private solicitor. This was indicative of a higher rate of legal representation than both the overall pattern in New South Wales where 37.7 per cent of persons charged are not represented by a lawyer,¹¹ and RR 4 where 30 per cent were not represented.

3.31 Although the proportion in this study is smaller, nevertheless there are important implications for those who are not legally represented, given the research reported in Chapter 1 indicating the possibility of more positive results for those accused persons who have legal representation.

ALCOHOL AND DRUG CONSUMPTION

3.32 An additional series of questions on alcohol and drug consumption was included in the survey form for this study. A total of 69 respondents (78.4 per cent of the whole sample, or 81.2 per cent of those who answered this question) indicated that they were drinking alcohol on the day of the alleged offence. Table 8 shows the amounts of alcohol indicated as having been consumed by those in the total sample who responded to the question, and those who fell below a standard score (SS) of 70 and who therefore could be categorised as intellectually disabled (see section below on Performance on the Kaufman Brief Intelligence Test, the K-BIT).

Table 8: Amount of alcohol consumed

Quantity of alcohol	% of total sample	% of ID sample
None	3.6	-
2-5 drinks	14.3	12.0
8-10 drinks	10.7	4.0
4-6 large bottles of beer	5.4	4.0
Half a carton of beer	7.1	12.0
1 carton of beer/1 cask	14.3	24.0
2 cartons	14.3	16.0
4 cartons	3.6	4.0
1 flagon + 2-5 cartons	8.9	12.0
Fair bit/too much	7.1	8.0
Lots/heaps	10.7	4.0
	100.0 N = 56	100.00 N = 25

3.33 There are no significant differences in alcohol consumption between the total group and the sub-sample with SS<70. This is an important finding because it indicates that people with intellectual disabilities are consuming as much alcohol as non-disabled people, prior to engaging in behaviour which brings them to the attention of the police.

3.34 One of the notable aspects of the findings reported in Table 8 is the sheer quantity of alcohol consumed. In answer to the question “Were you intoxicated at that time?” 79.4 per cent of those who responded indicated that they were intoxicated, and 20.6 per cent stated that they were not. These responses are slightly at odds with the next question which asked “What effect did the alcohol have on you?”, to which 13 per cent responded that they had not been affected by the alcohol, and a further 15.6% stated that they had been slightly affected. Respondents appeared to have some insight into the effect of alcohol, as most who reported drinking large quantities of alcohol also indicated that they felt drunk. Thus, most did not appear to be denying the effect of alcohol upon their state of soberness.

3.35 Ten participants (12.7 per cent) indicated that they had consumed drugs on the day of the offence, implying that alcohol is clearly the substance of choice in these two rural towns, at least amongst the offending population. One person stated that the drug consumed was ulcer medication, and one reported using pain killers, whereas the majority (62.5%) of those who indicated that they had used drugs indicated using marijuana. Five of those who had consumed marijuana had also consumed at least two cartons or more of alcohol. Polydrug use is an increasing problem, and this trend may have implications for treatment programmes for these accused persons.

PERFORMANCE ON THE K-BIT

3.36 Table 9 shows the numbers of participants falling into the various standard score categories on the K-BIT.

Table 9: K-BIT standard scores

Standard Scores*	Frequency	Percentage
100+	18	20.9
90-99	9	10.5
80-89	10	11.6
70-79	18	20.9
<70	31	36.0
	N = 86	99.9

*(Unless otherwise specified, these categories are used in K-BIT analyses throughout the report.)

3.37 The results show that 36.0 per cent of the sample (N=31) had a SS which was two standard deviations below the mean, that is below a score of 70, and could be regarded as falling into the intellectually disabled category. A further 20.9 per cent (N=18) obtained scores between 70 and 79, regarded as borderline. Thus a total of 56.9 per cent, more than half of the sample, have results indicating serious deficits in intellectual ability.

3.38 As in the previous study, social and adaptive skills were not assessed, owing to time constraints and privacy considerations since interviews were conducted in the relatively public environment of a court house. Inclusion of social and adaptive skills assessment may have indicated that some of the borderline group should definitely be regarded as mildly intellectually disabled, using the two-factor definition discussed in Chapter 1 of RR 4.

3.39 Tables 10 and 11 explore the lower end of the spectrum and show the distribution of scores below a standard score of 80; and the distribution of percentile ranks below 10.

Table 10: Distribution below standard score of 80

SS	Frequency	Cumulative %*
40	2	2.3
46	2	4.7
48	1	5.8
49	1	7.0
50	2	9.3
52	2	11.6
55	2	14.0
58	1	15.1
59	1	16.3
61	1	17.4
62	5	23.3
65	2	25.6
66	2	27.9
68	7	36.0
70	5	41.9
72	4	46.5
74	1	47.7
77	1	48.8
78	2	51.2
79	5	57.0
	N = 49

(*Cumulative percentage refers to the total sample.)

3.40 Six subjects obtained scores of less than 50, placing them in the moderately intellectually disabled category.

Table 11: Distribution of percentile ranks 10 and below.

Percentile rank	Frequency	Cumulative %*
<1	15	17.4
1	9	27.9
2	12	41.9
3	4	46.5
4	1	47.7
6	1	48.8
7	2	51.2
8	5	57.0
10	1	58.1
	N = 50

(*Cumulative percentage refers to the total sample.)

3.41 The seriousness of the cognitive deficits experienced by these low-functioning respondents is indicated by the fact that falling at or below a percentile rank (PR) of 10 means that 90 per cent of the general population is functioning better than these individuals.

3.42 The mean standard score for the total sample was 78, with a standard deviation of 18.9 and a minimum score of 40 and a maximum of more than 110. The mean PR was 19.4 with a standard deviation of 24.4, a minimum score of less than one and a maximum of 75.

3.43 These results indicate that overall the total sample is functioning much lower than the general population. By definition the mean standard score in the general population is 100, and a PR of 50 is average.

Table 12: Number of schools, and SS

SS	1	2	3	4	5+	N
100+	1	8	2	2	3	16
90-99	2	4	1	-	1	8
80-89	3	3	1	-	1	8
70-79	2	7	3	1	3	16
<70	10	4	7	-	2	23
	18	26	14	3	10	N = 71

3.44 There was a slight tendency for more of the respondents with SS=70-79 and <70 to report that they attended few schools, but there was no significant difference between those below 79 and those obtaining higher scores.

3.45 There were no significant differences in the proportion of males and females in the various standard score categories; nor were there significant differences with respect to the type of offence with which they were charged, or the number of previous offences.

3.46 The difficulty of accurately identifying an accused person with an intellectual disability is demonstrated by the fact that 18 of those with SS of less than 70, and 14 with SS=70-79 were classified as by the researchers as not having an intellectual disability. Three with SS=70-79 were fairly accurately described as “ID slight”. The accuracy at the other end of the scale is greater - no respondent with a standard score of over 80 was classified as being intellectually disabled by the researchers. Thus the greatest difficulty in identification of intellectual disability is the occurrence of false negatives, that is, people who are incorrectly classified as having no problem when in fact they have an intellectual disability. The researchers were not experts in the field of intellectual disability, and no criticism of them is intended - rather, these findings reflect the difficulty in identification of intellectual disability encountered by lawyers and other court and health professionals.

MINI-MENTAL STATE EXAMINATION (MSE)

3.47 Table 13 shows the cross tabulation between MSE and K-BIT scores, and also indicates the proportion of the sample falling into the various categories on the MSE, those having a score of below 22 being regarded as possibly having some type of mental abnormality and requiring further specialist assessment. The mean score on the MSE was 21.6 (standard deviation 4.1, minimum 10, maximum 26), not significantly lower than the RR 4 where the mean MSE was 22.5.

Table 13: MSE and K-BIT results

	K-BIT					MSE
MSE Value	100+	90-99	80-89	70-79	<70	MSE Frequency	MSE %	MSE Cum. %
22-26	16	8	8	9	10	51	59.3	59.3
19-21	2	1	1	5	7	16	18.6	77.9
<19	-	-	1	4	14	19	22.1	100.0
						N = 86	100.0

[Link to text only version of table]

3.48 Table 13 shows that 40.7 per cent fell below the cut off point of a score of less than 22, with 22.1 per cent being seriously affected (compared with 31 per cent, and 20 per cent respectively in RR 4).

3.49 On the other hand, 19 persons with SS<79 achieved satisfactory results in the MSE, whilst 5 with SS greater than 80 were unsatisfactory on the MSE and in need of further assessment.

3.50 Table 14 shows the number of males and females falling above and below the cut off mark.

Table 14: MSE and Gender

MSE Score	Males	Females	Total
22-26	43	8	51
19-21	12	4	16
<19	17	2	19
	72	14	86

(No significant differences were found.)

3.51 No significant differences were found between age categories and scores on the MSE, nor for category of most serious charge. Unemployment was significantly related to MSE score, however, (chi square 7.92, df=2, p<0.01). Of those respondents with a score of less than 19, 89.5 per cent were unemployed, whereas 71.4 per cent of those with a score of 19-21 and 54.0 per cent of those with a score of 22-26 were unemployed. Clearly, mental state has an interactive effect with employment.

3.52 Discriminant analyses were performed in an attempt to find variables which discriminated between those with SS<70 and those with SS>100. On the first analysis, the following variables were entered: work, pension, Serial Sevens, Backwards Spelling, Location, Orientation, Registration, and Three Stage Command. The following variables were found to be significant in discriminating between the groups at the p<.000 level of significance (degrees of freedom 4, 43) presented in descending order of strength: Serial Sevens subtest; being unemployed; being on a pension; and Location subtest.

3.53 The second discriminant analysis entered the variables of sex, work, drinking on the day of the offence, and total score on the MSE. The variables which were significant in discriminating between the groups at the p<.000 level of significance (degrees of freedom 2, 43) in descending order were: total MSE score, and drinking on the day of the offence.

3.54 These results indicate that those persons with lower standard scores are more likely to have lower levels of performance on Serial Sevens, Location and total score on the MSE, that is, to require further mental state examination; were less likely to be working, and more likely to be receiving a pension; and more likely to be drinking on the day of the offence.

CORRELATIONS BETWEEN SECTIONS OF THE SURVEY INSTRUMENT

3.55 The standard score (SS) obtained on the Kaufman Brief Intelligence Test (K-BIT) correlated significantly with the Mini-Mental State Examination (MSE) results (r=0.55, p<0.001) at a level almost identical to the results in RR 4 (r=0.51, p<0.000). The percentile rank (PR) on the K-BIT correlated with the total score on the MSE significantly also (r=0.47, p<0.001). As was the case with RR 4, these findings have important implications.

3.56 First, the significant agreement of the K-BIT SS and PR with the MSE indicates that to some extent they are assessing similar qualities. Secondly, the overlap or correlation between the two is not complete. Approximately 75 per cent of the variance is accounted for by other variables which might include English language proficiency (the MSE relies on English to a greater extent than the K-BIT, and the MSE cannot be entirely translated even if the subject has an interpreter present), and transient factors such as stress, anxiety, intoxication or mental illness.

3.57 The subsections of the MSE correlated with the K-BIT SS to the extent shown below:

MSE sub-section	K-BIT SS	K-BIT PR
Orientation	.34**	.32*
Location	.35**	.23
Registration	.05	.08
Serial sevens	.54**	.51**
Backwards spelling	.35**	.23
Three stage command	-0.9	-0.2
Total	.55*	.47**

(1-tailed test of significance, *p<.01, **p<.001)

3.58 These results differ from the results in the earlier court cohort study because Orientation and Location are now included in the group of subtests which have significant correlation with the K-BIT results. Serial Sevens emerges in the current study as having a higher correlation with K-BIT results, whereas Backwards Spelling has a slightly weaker correlation, which is not statistically significant.

3.59 These results indicate that the subsections of the MSE most reliant upon cognitive processes reflected by literacy and numeracy correlate most highly with the K-BIT. For this sample, however, questions pertaining to where the test is taking place, and the date and time, that is, short term memory function and awareness are also related to the subjects’ cognitive test results, whereas subtests reliant upon ability to register information in very short term memory in order to complete a new task are not related strongly to the K-BIT results.

THE COMBINED SAMPLE

Correlations between sections of the survey instrument

3.60 The standard score (SS) obtained on the K-BIT correlated significantly with the MSE results (r=0.54, p<0.001) for the total sample.

3.61 The sub-sections of the MSE correlated with the K-BIT SS to the following extent:

MSE sub-section	K-BIT SS
Orientation	.24*
Location	.09
Registration	-.047
Serial sevens	.52**
Backwards spelling	.42**
Three stage command	.02
TOTAL	.54**

(2-tailed test of significance, *p<0.01, **p<0.001)

3.62 Thus, for the total sample the sub-sections most strongly related to the test of cognitive ability, the K-BIT, are Serial Sevens and Backwards Spelling, followed by Orientation.

(a) Gender

3.63 When data from both court cohort studies are combined, a total of 208 cases are included. Males comprise 80.6 per cent (N=166) and females 19.4 per cent (N=40) of the sample, with no data being obtained on this variables for 2 cases.

(b) Ethnicity

3.64 Aboriginal persons comprised 34.8 per cent of the total sample (N=69), Islanders 2.5 per cent (N=5) and non-Aboriginal persons 62.6 per cent (N=124). The Aboriginal and Torres Strait Islander population for New South Wales in 1993 (the last census) was 70,020 persons, or 1.2 per cent of the population, a rise from 1.1 per cent in 1986.¹²

3.65 The level of over-representation of Aborigines in custody in New South Wales is 13.1, with 14.3 per cent of persons in custody being Aboriginal.¹³ Data pertaining to the proportion of Aborigines appearing before New South Wales local courts have not been located. These findings indicate that Aborigines are over-represented in this sample compared with the proportion in custody, and are vastly over-represented compared with the proportion of the New South Wales population. Nevertheless, it should be noted that these samples are not necessarily a random sample of local courts in New South Wales, and indeed the two courts most recently studied were selected because they were likely to have a high proportion of Aboriginal accused persons attending. The proportion of persons born overseas is lower in the total sample than in the New South Wales population; in 1991, 74.8 per cent of New South Wales residents were born in Australia, compared with 80.6 per cent of the sample.¹⁴ The over-representation of Aborigines clearly contributes to the higher proportion who were Australian born.

3.66 In the total sample, 61.3 per cent reported that they were unemployed, in comparison with 9.5 per cent of the New South Wales population in May 1994.¹⁵ The lowest unemployment rate is in the Northern, Far West, North Western and Central West regions, where the rate is 6.8 per cent compared with 8.5 per cent in Sydney (although there are considerable regional variations within metropolitan Sydney). Approximately 23 per cent of all people employed in New South Wales in May 1994 were part-time workers, compared with 36.1 per cent in this sample. In the sample, 59.3 per cent reported being in receipt of a social security pension or benefit.

PERFORMANCE ON THE K-BIT

Table 15: K-BIT standard scores for Non-Aboriginal, and Aboriginal/Islander respondents

	Non - Aboriginal		Aborigines/Islanders		Total
Score	Frequency	%	Frequency	%	Frequency	%
100+	57	44.5	10	14.1	67	33.7
90-99	30	23.4	5	7.0	35	17.6
80-89	13	10.1	9	12.6	22	11.1
70-79	11	8.6	17	23.9	28	14.1
<70	17	13.3	30	42.3	47	23.6
	N = 128	99.9	N = 71	99.9	199	100.1

[Link to text only version of table]

3.67 These results show that 23.6 per cent of the total sample (N=47) had a SS which was two standard deviations below the mean, that is, below a score of 70. This group would therefore be regarded as falling within the category of intellectually disabled. A further 14.1 per cent has SS=70-79, in the borderline range of intellectual ability. A total of 37.7 per cent of the sample, more than one third, obtained results on a non-verbal, untimed test of cognitive ability which indicated serous deficits in cognitive skills.

3.68 No significant differences were found between the standard score categories and the type of offence charged. Assault was the most frequent offence in all standard score categories.

3.69 There is a significant difference between the Aboriginal/Islander results on the K-BIT, compared with non-Aborigines (Chi square=50.22, df=8, p<0.000). More non-Aborigines scored 100 or above, and fewer scored less than 79.

3.70 Table 16 shows the relationship between the K-BIT and the scores on the MSE for the total sample.

Table 16: Cross tabulation of K-BIT results with MSE

	<70	70-79	80-89	90-99	100+	Total
22-26	15	9	13	26	46	109
	8.9	5.3	7.7	15.4	27.2	64.4
19-21	12	9	2	3	3	29
	7.1	5.3	1.2	1.8	1.8	17.2
<19	18	6	9	-	4	31
	10.6	3.6	1.8	-	2.4	18.4
	45	24	18	29	53	169
	26.6	14.2	10.7	17.2	31.4	100.0

3.71 The chi square test is significant, at the p<0.000 level of significance (df=8).

3.72 The table indicates that whilst there is some clustering of low MSE with low K-BIT SS, and high with high, nevertheless the overlap is not total. For example, 15 people with a SS of less than 70 obtained a score on the MSE which would indicate no further need for mental status examination. At the other end of the spectrum, four people with MSE score of less than 19 scored over 100 on the K-BIT. These results indicate that the MSE cannot be substituted for a test of general intelligence.

3.73 Discriminant analysis was performed in an attempt to find variables which distinguished between those subjects with SS<70 and SS>100, entering the variables Orientation, Location, registration, Serial Sevens, Backwards Spelling, Three Stage Command, Aboriginality, work, and pension. The variables which distinguished most strongly between the groups at the p>.000 level (degrees of freedom 4, 83) were: Serial Sevens, Aboriginality, work, and Backwards spelling. Thus, low scorers on the K-BIT were more likely to score poorly on Serial Sevens, be Aboriginal or Islander, be unemployed and score poorly on Backwards Spelling.

3.74 Using a slightly different approach, discriminant analysis was performed in an attempt to distinguish between Aboriginal and non-Aboriginal people appearing before these local courts, entering the variables Orientation, Location, Registration, Serial Sevens, Backwards Spelling, Three Stage Command, Total MSE score and SS. The variables which were found to discriminate most strongly at the p<.000 level (degrees of freedom 4, 154) in descending order were: Standard Score, Location, Orientation and Backwards Spelling. Aboriginal/Islander respondents were more likely than non-Aborigines to perform poorly on the K-BIT and on the Orientation subtest, but to do well on Location and Backwards Spelling.

3.75 The Aboriginal and Islander subgroups were combined and a discriminant analysis performed to investigate within this population which variables discriminated between those with SS>100, and SS<70. The variables entered into the equation were Orientation, Location, Registration, Serial Sevens, Backwards Spelling, Three Stage Command, work and pension. The two variables found to discriminate at the p<.005 (degrees of freedom 2, 35) were in descending order, Serial Sevens and Location. These results, related to the MSE assessment, may indicate that those Aboriginal participants who had multiple problems, such as both intellectual disability and other mental abnormality, may be functioning overall at a lower level on tasks requiring cognitive reasoning.

3.76 Discriminant analysis performed on gender groups entered the variables age, Aboriginality, work, pension, total MSE score and SS on the K-BIT. The variables which distinguished between the groups at the p<.01 or greater (degrees of freedom 3, 141) were in descending order: receiving a pension, working, total MSE score, and age. Women were more likely not to be receiving a pension, yet to be unemployed, to have a low MSE score, and to be older than male accused persons.

3.77 Results of these discriminant analyses, however, must be regarded with caution. First, not all variables could be entered into the equation because many were not continuous variables. Secondly, variables which had high levels of missing data could not be entered because the remaining cases on which to perform the analyses were too few.

1.New South Wales Law Reform Commission People with an Intellectual Disability and the Criminal Justice System: Appearances Before Local Courts (Research Report 4, 1993).

2. NSW Bureau of Crime Statistics and Research New South Wales Criminal Court Statistics 1994 (Attorney General’s Department, 1995).

3. J Powles and S Gifford “How Healthy are Australia’s Immigrants?” in J Reid and P Trompf (eds) The Health of Immigrant Australia - A Social Perspective (Harcourt Brace Jovanovich, Sydney, 1990) at 77-107; New South Wales Year Book 1995 (No 75, Australian Bureau of Statistics, Sydney, 1995).

4. New South Wales Year Book 1995, at 128ff. The Australian Bureau of Statistics definition includes people aged 15 years or over who are not employed, who are available for work, and who are taking active steps to find work.

5. Department of Health, Housing, Local Government and Community Services Regional Labour Market Disadvantage, Social Justice Research Program into Locational Disadvantage (Report 16, AGPS, 1993) Appendix 1.

6. Australian Bureau of Statistics telephone information service, October 1995.

7. New South Wales Year Book 1995.

8. I Castles Australian National Classification of Offences (Australian Bureau of Statistics, 1986).

9. NSW Bureau of Crime Statistics and Research (1995) Table 1.1 at 8-9.

10. S C Hayes and G Craddock Simply Criminal (2nd ed, Federation Press, Sydney, 1992).

11. NSW Bureau of Crime Statistics and Research (1995) Table 1.4 at 12.

12. New South Wales Year Book 1995.

13. D Biles and D McDonald (eds) Deaths in Custody Australia, 1980-1989 (Australian Institute of Criminology, Canberra, 1992).

14. New South Wales Year Book 1995.

15. New South Wales Year Book 1995.