Test of Grocery Shopping Skills

Schizophrenia: (Hamera & Brown, 2000, n = 26, Mean Age = 40.5 (range 22 - 57) years)

• Adequate test-retest reliability between forms 1 and 2 of TOGSS (0.64 ≤ r ≤ 0.84 across all scales) administered over two different testing periods in a test-retest design

Interrater Reliability:

Schizophrenia

• Excellent Intercoder reliability assessed between two independent raters (85.7%) for responses during semi-structured interviews (Hamera & Brown, 2000)
• Excellent interrater reliability for TOGSS for both forms 1 (r = 0.99) and 2 (r = 1.00) (Hamera & Brown, 2000)

Predictive Validity:

Serious Mental Illness(Schizophrenia-spectrum, bipolar disorder, or other depression/anxiety diagnoses and accompanying functional disability): (Faith & Rempfer, 2018, n = 59; Mean Age = 43.19 (9.72) years; Of these, n = 21 schizophrenia or schizoaffective disorder, n = 15 bipolar disorder, n = 23 other depressive and/or anxiety disorders)

TOGSS accuracy scores provide incremental predictive value that self-report alone cannot provide in predicting real-world shopping skills (R2 change = 0.07, F (1,49) = 4.37, p < 0.05)

Convergent validity:

Schizophrenia:

(Greenwood, 2005; n = 75; Of these, n= 28 with negative symptoms, Mean Age = 35.1 (8.8) years, n=25 without negative symptoms, Mean Age = 35.3 (10.6), n= 22 healthy control, Mean Age = 36.2 (12.9))

• TOGGS accuracy was significantly correlated to Key Search Task for Spatial Strategies (z = 2.48, p < 0.05)
• TOGGS accuracy was significantly correlated to Letter Number Span Task for Working Memory (z = 0.48, p < 0.001) 

• TOGGS redundancy was significantly correlated to Phonological Fluency Test for Response Initiation (z = 4.02, p < 0.01)

(Hamera & Brown, 2000)

• Excellent convergent validity between forms 1 and 2 of TOGSS (r ≥ 0.60 across all scales)
• Excellent convergent validity between TOGSS and Drug Store Shopping Skills (r ≥ 0.65 across subscales, p < 0.001)

(Rempfer et al., 2003, n = 73; Mean Age = 39.46 (9.9) years)

• Adequate convergent validity between redundancy and increased time (r = 0.35, p < 0.01).
• Adequate correlation between TOGGS redundancy and Wisconsin Card Sorting Task (WCST) total (r = - 0.39, p<0.01)
• Adequate correlation between TOGGS redundancy with Rey Auditory Verbal Learning Test (RVLT) verbal recall total (r = -0.34, p<0.01)
• Adequate correlation between TOGGS accuracy and Stroop word trial (r = 0.32, p<0.05)
• Poor correlations between TOGGS accuracy (r = 0.29, p<0.01) and redundancy (r = -0.20, p<0.05) with Stroop color trial
• Poor correlation between TOGGS redundancy and FAS verbal fluency (r = -0.27, p<0.05)
• Poor correlation between TOGGS redundancy and Stroop color-word trial (r = -0.26, p<0.05)
• Poor correlation between TOGGS time and RVLT learning (r = 0.24, p<0.05)
• Poor correlation between TOGGS redundancy and Allen Cognitive Levels Test (ACL) letter cancellation time (r = 0.24, p<0.05)
• Poor correlations between TOGGS accuracy (r = - 0.23) and redundancy (r = 0.23) with WCIST perseverative errors (p<0.05)

(Rempfer, McDowd, & Brown, 2017, n = 81 Mean Age = 41.27 (8.6) years)

• Adequate correlations comparing the TOGGS with the learning potential measure WCST in the Post training group (r = .305, p < 0.01)
• Poor correlations comparing the TOGGS with the learning potential measure Rey Osterrieth Complex Figure Test (ROCFT) in the Post training group (r = .259, p < 0.05)
• Poor correlations comparing the TOGGS with the learning potential measure WCST in the Categorical group (tau b = .209, p < 0.05)
• Poor correlations comparing the TOGGS with the learning potential measure ROCFT in the Post training group (𝝩𝛃 = .234, p < 0.05)
• Poor correlations comparing the TOGGS with the learning potential measure WCST in Regression residual group ( 𝝩𝛃 = .247, p < 0.05)

(Zayat et al., 2011, n = 80; Mean Age = 42.67 (8.47) years)

• Adequate correlations comparing the TOGGS redundancy outcome with the neurocognitive measure Wisconsin Card Sorting Task (WCST) perseverative error (r = .382, p < 0.01)
• Adequate correlations comparing the TOGGS redundancy outcome with the neurocognitive measure WCST (r = .323, p < 0.01)
• Adequate correlations comparing the TOGGS redundancy outcome with the neurocognitive measures WCST Total Correct (r = -.313, p < 0.05)
• Adequate correlations comparing the TOGGS redundancy outcome with the neurocognitive measure Letter Numbering Sequence (LNS) (r = -.323, p < 0.01)
• Adequate correlations comparing the TOGGS accuracy outcome with the neurocognitive measure Trail Making Test (TMT) (r = -.325, p < 0.05)

• Poor correlations comparing TOGGS accuracy to the neurocognitive measure: WCST Total Correct (r=.222, p < 0.05)
• Adequate correlations comparing the TOGGS accuracy outcome with the neurocognitive measure LNS (r = -.385, p < 0.01)
• Adequate correlations comparing the TOGGS accuracy outcome with the neurocognitive measure Digital Span Backward (DSB) (r = -.339, p < 0.01)
• Adequate correlations comparing the TOGGS accuracy outcome with the neurocognitive measure Months Ordering Test of Working Memory (MO) (r = .422, p < 0.01)

Discriminant Validity:

Schizophrenia:

• Excellent discriminant validity, where there were no significant relations between outcomes of accuracy and redundancy (r = -0.22), or accuracy and time (r = -0.09). (Rempfer et al., 2003)

Serious Mental Illness:

(Faith & Rempfer, 2018)

• Adequate construct validity of the TOGSS regarding the accuracy outcome was significantly associated with both directly observed real-world shopping skills (r = 0.424, p < 0.01) and self-reported shopping skills (r = 0.444, p < 0.01) 
• Naturalistic behavioral observation in real world settings is considered the best estimation of a person's real life functioning in absence of a gold standard for assessing real life functioning

(Hamera, Brown, Rempfer, & Davis, 2002, n = 64; Of these, n = 26 schizophrenia, Mean Age = 44.38 (11.19) years; n=19 bipolar disorder, Mean Age = 41.32 (9.32) years;  n = 19 no diagnosed mental illness, Mean Age = 34.63 (9.29) years)

• Adequate construct validity for TOGGS accuracy:  Age is negatively correlated with accuracy in the group with schizophrenia and the group with bipolar disorder (r = -.36 and -.58, respectively) and positively correlated in the group without mental illness (r = .46) 
• TOGGS redundancy:  Significant differences between the group with mental illness compared to the group without mental illness (F = 5.30, (2,60), p < .01)
• TOGGS time: The group with mental illness required significantly more time compared to the group without mental illness (F = 6.14, (2,60), p < .01)

(Rempfer & Fowler, 2018, n = 49; Mean Age = 45.73 (10.69 years);  Of these, n = 22 schizophrenia, n = 7 schizoaffective disorder, n = 14 major depressive disorder, n = 6 bipolar disorder)

• Excellent correlations with the neurocognitive measure California Verbal Learning Test-II (CVLT-II) (r = .62, p < 0.01)
• Adequate correlations with the neurocognitive measure Knowledge of Grocery Shopping Skills (KOGSS) (r = .43, p < 0.01)
• Adequate correlations with the neurocognitive measure UCSD Performance-based Skill Assessment (UPSA-B) (r = .57, p < 0.01)
• Adequate correlations with the neurocognitive measure  Controlled Oral Word Association Test (COWAT) (r = .38, p < 0.05)
• Adequate correlations with the neurocognitive measure  d2 Test of Attention (D2) (r = .53, p < 0.01)
• Adequate correlations with the neurocognitive measure LNS (r = .44, p < 0.01)
• Adequate correlation with the neurocognitive measure Trail Making Test, part B (Trails B) (r = .41  p < 0.01)