METHODOLOGY
TABLE 4.15 TABLE OF COMORBID PROBLEMS ADHD
4.7 COGNITIVE TESTING
Three tests that are known to be sensitive to measuring attention/concentration and impulsiveness in childhood were conducted on participants, the Continuous Performance Test, the Matching Familiar Figures Test and the Letter Cancellation Test. An assessment of time estimation, both prospective and retrospective, was also made. The data are presented in Table 4.16. After correcting for the effect of multiple testing, two results were lost (relating to prospective time estimation and letter cancellation). All other results remained significant at « < .05 level.
4.7.1 Continuous Performance Test
The Continuous Performance Test [CPT] is a computerised test of sustained vigilance in a signal detection task. Each stimulus picture is presented for one second with a one second interstimulus interval. Altogether 268 stimuli are presented, among which there are 32 pairs o f successively identical stimuli. The computer records the number o f correct identifications (hits), incorrect identifications (false positives), failures to identify (false negatives) and the sum of the reaction times (RT) for hits and false
positives. From these results four scores were calculated: hits, false positives, false negatives and RT hits.
One-way analysis o f variance with the CPT reaction time as the dependent variable showed a significant main effect was present (F(2,85) = 10.36, p<.001). There was only one Bonferroni pairwise significant difference (« = .05) - the clinic control group took significantly longer than the normal control group (p< 001 ). There were no other
significant differences.
The number o f correct responses were recorded as the number o f hits. One-way analysis o f variance with the hits as the dependent variable showed a significant main effect was present (F(2,85) = 7.76, p=.001). There was only one Bonferroni pairwise significant difference (« = .05) - the ADHD group had fewer hits than the normal control group (p=.001). There were no other significant differences.
False negatives represent failures to identify a target (i.e. errors of omission). One-way analysis of variance with the false negatives as the dependent variable showed a significant main effect was present (F(2,85) = 7.76, p=.001). There was only one Bonferroni pairwise significant difference (« = .05) - the ADHD group made significantly more false negatives (or misses) than the normal control group (p=.001). There were no other significant differences.
False positives represent incorrect identifications (i.e. errors o f commission). One-way analysis o f variance with the false positives as the dependent variable showed a significant main effect was present (F(2,85) = 3.80, p=.03). There was only one Bonferroni pairwise significant difference = .05) - the ADHD group made significantly more false positives (or misfires) than the normal control group (p=.03).
There were no other significant differences.
These results suggest that the record of errors is more helpfijl than mean reaction time for this CPT assessment o f deficits associated with ADHD in adulthood. Mean reaction time discriminated between clinic controls and normal controls only, showing that clinic controls take longer to process incoming information than a normal control group. With respect to errors of commission and omission, the ADHD group consistently performed worse than the normal control group and they identified fewer hits overall. Thus when taking a similar time to process information as normal controls, the ADHD group make far more errors of omission and commission.
4.7.2 Matching Familiar Figures Test
This is a test o f impulsiveness vs reflectiveness in cognitive style. Impulsiveness is represented by an enduring disposition to respond rapidly but incorrectly in a situation where there is uncertainty about which response is correct. The task is to match a single picture with the identical member of an initial set. Three measures were taken;
the number o f times an individual gets it wrong (i.e. errors), the number that are identified correctly the first time, and the time that is taken to do each trial. The measure o f impulsiveness is the number of rapid errors.
One-way analysis of variance with the MFF RT as the dependent variable showed there was no significant main effect (F(2,87) = 2.88, NS). The number o f correct responses identified the first time was compared in one-way analysis of variance with the correct responses as the dependent variable. This showed a significant main effect was present (F(2,87) = 7.98, p=.001). Bonferroni post hoc pairwise comparisons (« = .05) revealed the ADHD group was significantly slower than the normal control group (p=.01) and the clinic control group (p=.001). There was no significant difference between scores for the normal control group and the clinic control group.
The number o f errors were also compared in one-way analysis o f variance with the errors as the dependent variable. This showed a significant main effect was present (F(2,87) = 10.72, p<.001). Bonferroni post hoc pairwise comparisons (« = .05) revealed the ADHD group made significantly more errors than the normal control group (p=.002) and the clinic control group (p< 001). There was no significant difference between scores for the normal control group and the clinic control group.
Thus a similar pattern of results was found in analysing the MFF data and the CPT data. Mean reaction time did not discriminate between groups, but variables relating
to the number o f errors and the number identified correctly the first time revealed that the ADHD group made significantly more errors and identified fewer pictures correctly first time than both clinic and normal control groups.
4.7.3 Letter Cancellation Test
The Letter Cancellation Test [LCT] is a measure o f ability to sustain attention. This consists o f rows o f letters randomly interspersed with a designated target letter and performance is scored for errors and number of lines completed within a two minute allocated time.
The association between the number of lines completed within the time period and the number o f errors made was examined in one-way multivariate analysis of variance with the number o f errors and lines as dependent variables. There was a significant main effect that revealed differences between the groups (Wilks’ Lambda =.88; Exact F(4,170) = 2.68, p=.03). As a follow-up to this multivariate ANOVA, two one-way ANO VAS were carried out in order to isolate the source o f the differences between the groups. These analyses indicated that the source o f the difference stemmed from the letter cancellations (F(2,88) = 3.78, p<.03) rather than the number o f lines completed (F(2,88) = 1.48, NS). However, strictly speaking, these results for the LCT are not significant, given the number of tests carried out and the result was lost when the Bonferroni correction for multiple testing was applied.
TABLE 4.16 TABLE OF COGNITIVE TEST RESULTS