Page 58 - JCBP-3-2

Journal of Clinical and
Basic Psychosomatics Interpersonal relationship rating scale

The Tucker–Lewis Index (TLI), another fit index, 3.2.2. Test–retest reliability
also ranges from 0 to 1, with higher values (closer to Eighty-four participants were randomly selected from
1) reflecting a better fit. A TLI value >0.90 is typically Sample 2 (n = 335), and the Interpersonal Relationship
2
considered indicative of a good model fit. However, some Scale was administered again 2 months after the first
21
less stringent standards suggest that values of GFI, CFI, measurement. The results showed that the correlation
and TLI >0.80 are acceptable. coefficient of the total score between the two measures
The root mean square error of approximation (RMSEA) was 0.401. The correlation coefficients for the scores of the
assesses model fit by measuring how well a model dimensions between the two measures ranged from 0.269
approximates the data; lower values (closer to 0) indicate a to 0.381 (Table 3).
better fit. The p of close fit (PCLOSE) is a one-sided test of 3.2.3. Construct validity
the null hypothesis that RMSEA = 0.05, representing a close-
fitting model. In addition, the PCLOSE statistic provides Correlation analyses were conducted between the
22
the probability that the RMSEA is <0.05, indicating a close fit. dimensions of the scale as well as between the dimensions
and the total scale for both samples. The results showed that
In our study, the indices were as follows: GFI = 0.831
(>0.80), CFI = 0.860 (>0.80), TLI = 0.838 (>0.80), and the correlation coefficients between the four factors ranged
from 0.465 to 0.693 (all P < 0.01), and the correlation
RMSEA = 0.086 (<0.10). These results from the confirmatory coefficients between the factors and the total scale ranged
factor analysis indicate that the model fit for the Interpersonal from 0.792 to 0.881 (all P < 0.01). These results, detailed
Relationship Rating Scale is generally acceptable (Table 2).
in Table 4, indicate that the factors of the scale were
3.2. Reliability and validity analysis of the formal significantly correlated.
questionnaire 3.2.4. Criterion-related validity

3.2.1. Internal consistency reliability The Interpersonal Relationship Comprehensive Diagnostic
Internal consistency reliability was analyzed for Samples 1 Scale, with an alpha coefficient of 0.900 in the sample of
and 2 as well as the total sample. The total scale consisted this study (n = 882), was used as the validity criterion.
of 18 items, with an overall Cronbach’s alpha coefficient of Pearson’s product-difference correlation analysis was
0.903. The alpha coefficients for the four dimensions across performed to assess the correlation between this scale and
both samples and the total sample, as well as the total scale, the Interpersonal Relationship Rating Scale.
ranged from 0.674 to 0.909 (Table 3), indicating good The results of correlation analysis showed that the
internal consistency reliability. correlation coefficients between the factors of the two

Table 3. Cronbach’s alpha coefficients and test–retest reliability for the total scale and dimensions

Total scale Factor 1 Factor2 Factor 3 Factor 4
Sample 1 (n =547) alpha coefficient 0.899 0.806 0.819 0.711 0.674
1
Sample 2 (n =335) alpha coefficient 0.909 0.775 0.805 0.744 0.741
2
Total sample alpha coefficient 0.903 0.794 0.815 0.718 0.701
Test-retest reliability 0.401** 0.312** 0.344** 0.381** 0.269*
Notes: *P<0.05, **P<0.01.

Table 4. Correlation coefficients for dimensions and total scale
Sample 1 (n =547) Sample 2 (n =335)
1 2
F1 F2 F3 F4 F1 F2 F3 F4
F1
F2 0.465** 0.539**
F3 0.527** 0.605** 0.606** 0.693**
F4 0.593** 0.681** 0.662** 0.681** 0.684** 0.680**
Total 0.836** 0.792** 0.820** 0.862** 0.869** 0.815** 0.853** 0.881**
Note: **P<0.01.

Volume 3 Issue 2 (2025) 52 doi: 10.36922/jcbp.3625