Norms of the geriatric complex figure test for adults aged 65 and older in Shanxi Province: analyses of reliability and validity

The complexity of the graphics in the Rey Osterrieth Complex Figure (ROCF) imposes a cognitive load on older adults, leading to the development of the simplified Geriatric Complex Figure Test (GCFT), which, however, has yet to be tested or standardized. This study aimed to establish normative data for the GCFT among adults aged 65 and older in Shanxi Province and to evaluate its reliability and validity. This study adopted a cross-sectional design. A multistage, stratified, randomized cluster sampling method was used to select five communities in Shanxi Province, targeting residents aged 65 years and older. SPSS 26.0 and JASP 0.19.1.0 were used for norm development, item analysis, reliability assessment, exploratory factor analysis (EFA), and confirmatory factor analysis (CFA). The Mini-Mental State Examination (MMSE) was employed to assess criterion validity. Mean and percentile norms were established across subgroups based on gender, age, education level, and residential status (urban vs. rural), with no significant differences observed. The correlations between the three GCFT subtests—copying, 3-min recall, and 20-min recall—and the total score were 0.738, 0.929, and 0.920, respectively. All item-total correlations were positive and statistically significant (p < 0.001). The GCFT demonstrated good internal consistency (Cronbach’s alpha = 0.823) and split-half reliability (coefficient = 0.806). The EFA yielded a KMO value of 0.749, and Bartlett’s test of sphericity was significant (χ2 = 3707.715, df = 351, p < 0.001), resulting in nine factors explaining 64.7% of the cumulative variance. The CFA indicated a good model fit (χ2/df = 1.93, CFI = 0.923, TLI = 0.906, RMSEA = 0.047). Criterion validity analysis indicated a significant positive correlation between the GCFT and MMSE scores (r_partial = 0.429, p < 0.001). The established GCFT norms for adults aged 65 and above are representative and demonstrate good reliability and validity for assessing visuospatial and memory functions in this population.