2. MARCO TEÓRICO
2.6. ANÁLISIS DE LOS MÉTODOS QUE SE PODRÍA USAR EN LA
After checking the data for normality and outliers, two subjects’ data were removed as mentioned above. The RTs and MTs for the two no-go tasks were analysed separately using repeated measures analyses of variance (ANOVA). Large variance in the data of patient groups is a common finding in schizophrenia (Straube and Oades, 1992). In light o f this, the relatively small sample sizes, and the nature of the group differences that were of primary interest, it was decided to compare only two groups at a time. For
vs the ‘low symptom’ patients, the ‘high symptom’ patients vs the controls and the ‘low symptom’ patients vs the controls. The between subjects factor was Group (‘high symptom’ patients vs ‘low symptom’ patients, etc) and the within subject factors were Run (Run 1, Run 2, Run 3) for the SRT condition for Study 4, and Condition (CRTl, CRT2 or CRTS) for the CRT condition for Tasks 1 and 2. A one way ANOVA was used to analyse the data from the SRT condition of Study 5, with Group (‘high symptom’ patients vs ‘low symptom’ patients, etc) as the between subjects factor. In order to investigate the difference between increased choice complexity of target go stimuli and increased dimensional overlap o f target go and no-go stimuli an ANOVA was run using the CRT data from both tasks. The three subject groups were used together in one analysis in order to determine the trend across all groups. The between subjects factor was group (controls, ‘high symptom’ patients, and ‘low symptom’ patients) and the between subjects factors were Task (Study 4 vs Study 5) and Condition (CRTl, CRT2 or CRTS). In order to investigate the difference between SRT conditions using stimulus invariance or stimulus variability across trials the mean SRTs of Study 4 and the SRTs of Study 5 were compared using an ANOVA. The between subjects factor was Group (controls, ‘high symptom’ patients, and ‘low symptom’ patients) and the within subjects factor was Task (Study 4 vs Study 5). Where necessary, to deal with violations of assumptions of sphericity, the Greenhouse-Geisser epsilon was used to adjust the degrees of freedom. Significant main effects or group interactions were followed up with paired or independent t -tests, as appropriate. Due to computer error, the data of one ‘low symptom’ patient is missing for the second run of the SRT for Study 4. One way ANOVAs were used to examine the group differences in the cognitive tests except on tests with different conditions (Hayling Test). For this, a repeated measures ANOVA was used with Group as the between subjects factor and Condition (initiation or suppression) as the within subjects factor.
5.3 Results
The two groups of patients and the normals did not differ in terms of male to female ratio (x^ = 0.49, df = 2, p > 0.05), age [F(2,25) = 1.48, p > 0.05] handedness, [F(2, 25) = 2.66, p > 0.05] (Oldfield, 1971) or estimates of ‘premorbid’ verbal IQ obtained from the National Adult Reading Test (NART, Nelson and Willison, 1991) [F (2,25) = 0.46, p > 0.05). The two groups did differ in terms of scores on the Beck Depression Inventory (BDI, Beck et al., 1961) [F (2, 15) = 4.032, p < 0.05]. Post hoc Tukey’s test showed that this difference was caused by higher depression scores for the ‘high symptom’ patients than the controls (p < 0.05) but no other groups differed significantly. The two patient groups did not differ in terms of chronicity, dose of neuroleptic medication or dose o f anticholinergic medication (p > 0.05).
5.3.1 Study 4
5.3.1.1 Movement Time
A series of repeated measures ANOVA were carried out for the movement time (MT) data. Group (‘high symptom’ vs ‘low symptom’ patients, or ‘high symptom’ patients vs controls, or ‘low symptom’ patients vs controls) was the between subjects variable and Run (1, 2 or 3; SRT ) or Condition (CRTl, CRT2 or CRT3) were the within subjects variables. The mean MTs and the results of the ANOVAS are presented in Table 5.3. There were no significant main effects of Run in the SRT ANOVAs or Condition in the CRT ANOVAS. There were no significant Group by Run or Group by Condition interactions in any of the ANOVAs for MTs. The ‘high symptom’ and ‘low symptom’ groups showed no significant differences in MT. Each patient group had significantly slower MTs for the SRT and CRT conditions than the control group (p < 0.05).
Table 5.3. The mean MTs and standard deviation (SD) for each group for Study 4 and the results of the ANOVAs on MTs. Where necessary, to deal with violations of assumptions of sphericity, the Greenhouse-Geisser epsilon was used to adjust the
degrees of freedom. *(p < 0.05)
Group Mean MT SRT (sd) Mean MT CRT (sd)
High patients 244.38 (52.43) 266.76 (55.16) Low patients 255.89 (83.27) 276.19 (80.41)
Controls 196.28 (38.43) 192.63 (34.23)
Condition Source DF F value P value
High vs Low SRT Run 1,14 0.07 0.85
Symptom
Patients Group X Run 1,14 0.31 0.64
Group 1,11 0.09 0.77
CRT Condition 2, 24 0.28 0.76
Group X Condition 2, 24 1.64 0.21
Group 1,12 0.07 0.80
High Symptom SRT Run 1,25 0.09 0.86
Patients vs
Controls Group X Run 1,25 0.07 0.88
Group 1,17 5.31 0.03*
CRT Condition 2, 34 1.18 0.32
Group X Condition 2, 34 1.87 0.17
Group 1,17 13.26 0.002*
Low Symptom SRT Run 2, 32 0.90 0.42
Patients vs
Controls Group X Run 2, 32 1.44 0.25
Group 1, 16 4.47 0.05*
CRT Condition 2, 34 1.34 0.28
Group X Condition 2,34 0.74 0.49
Group 1,17 10.15 0.005*
5.3.1.2 Error Data
Very few errors of any type were made by the patients or normals. The median error data are shown in Table 5.4. A series of Kruskal-Wallis tests revealed no significant differences among the three groups in the number of anticipations, decision errors, or long responses in the SRT condition (p > 0.05). In the CRT condition there was a significant group difference in the number of errors of commission (going on a no-go)
(p = 0.02) and the number of anticipation errors (p = 0.02). Further analyses of these effects revealed that the patients in the ‘high symptom’ group made significantly more
Table 5.4. Median error values (and range) for Study 4 and Study 5. Values in marked boxes for the controls and the ‘low symptom’ patients differ significantly from the ‘high symptom’ patients (p < 0.05)
Anticipation Long Omission Commission Partial
Responses Errors Study 4 SRT Controls 0.00 0.00 0.00 / / (0.00- 0.67) (0.00- 0.33) (0.00- 0.33) High 0.00 0.00 0.00 / / Patients ( 0 .0 0 -0 .3 3 ) (0.00 - 0.00) (0.00-0.00) Low 0.00 0.00 0.00 Patients ( 0 .0 0 -1 .0 0 ) (0.00 - 0.33) (0.00-0.00) CRT Controls 0.00 0.08 0.00 0.67 8.17 (0.00- 0.67) (0.00 - 0.67) (0 .0 0 - 1.67) (0.00 - 2.00) (0.67 - 33.67) High 0.33 0.00 0.33 1.67 8.00 Patients (0.00-0.67) (0.00- 0.33) (0.00- 0.67) (0.00 - 7.67) (0 .0 0 -1 2 .6 7 ) Low 0.00 0.00 0.00 0.00 6.00 Patients ( 0 .0 0 -0 .3 3 ) ( 0 .0 0 - 1.67) ( 0 .0 0 - 1.33) ( 0 .0 0 - 1.00) (0 .0 0 -1 5 .6 7 ) Study 5 SRT Controls 0.00 0.00 0.00 / / (0.00- 3.00) (0.00- 0.00) (0.00- 0.00) High 0.00 0.00 0.00 / / Patients (0.00 - 6.00) (0 .0 0 - 1.00) (0.00-0.00) Low 0.00 0.00 0.00 Patients ( 0 .0 0 -1 .0 0 ) (0.00 - 0.00) (0.00-0.00) CRT Controls 0.00 0.00 0.00 0.00 1.17 (0.00- 0.33) (0.00 - 4.33) (0.00 - 0.33) (0.00 - 0.00) (0.00 - 8.00) High 0.00 0.00 0.00 0.00 1.33 Patients (0.00-0.00) (0.00- 0.33) (0.00- 0.33) (0.00 - 2.00) ( 0 .6 7 -2 .3 3 ) Low 0.00 0.00 0.00 0.00 0.67 Patients (0.00 - 0.67) ( 0 .0 0 -0 .0 0 ) (0 .0 0 - 0 .6 7 ) (0 .0 0 - 0 .3 3 ) (0.33 - 6.00)
errors of commission than the ‘low symptom’ group (p = 0.007) and the normal controls (p = 0.05). The ‘high symptom’ group also made significantly more anticipation errors compared to the normal controls (p = 0.02) but this comparison did not reach significance between the two patient groups. There were no significant differences among the three groups in terms of the number of errors of omission, long responses or partial errors (p > 0.05).
5.3.1.3 SRT condition
The mean RTs for the SRT condition of each group are presented in Figure 5.3. An ANOVA examined the difference between the ‘high symptom’ patients and controls. The main effect of Group was significant [F(l,17) = 7.44, p = 0.01] with the ‘high symptom’ patients having slower RTs (406.62, sd = 82.49) than the controls (336.81, sd = 54.34). The main effect of Run was significant. Further investigation of this effect revealed that the RTs for Run 2 (359.56 ms, sd - 63.04) were significantly slower than those for Run 1(330.58 ms, sd = 43.71) (t =3.09, df = 18, p = 0.006) but no other Runs differed significantly. The Group x Run interaction was not significant.
In the ANOVA between the ‘low symptom’ patients and the controls the main effect of Group was not significant (p > 0.05) but the main effect of Run was significant [F(l, 21) = 4.81, p = 0.03]. Further analysis of this effect revealed that the RTs for Run 2 (359.56 ms, sd = 63.04) were significantly slower than the RTs for Run 1 (330.58 ms, sd = 43.70) but no other Runs differed significantly. The Group x Run interaction was not significant.