Page 128 - Vseživljenjsko učenje kot temelj trajnostne družbe
P. 128
Vita Jeraj and Petra Dolenc
Results
Psychometric Properties of the Slovenian Version of the TEIQue-SF
The assumed 4-factor structure of the TEIQue-SF questionnaire was evalu-
ated using confirmatory factor analysis (CFA) with maximum likelihood es-
timation. The model fit was assessed using several fit indices, including the
2
ratio of chi-square to degrees of freedom (χ /df), the Comparative Fit Index
(CFI), the Tucker-Lewis Index (TLI), the Root Mean Square Error of Approxi-
mation (RMSEA) with a 90 confidence interval, and the Standardized Root
Mean Square Residual (SRMR). The following thresholds were used to deter-
2
mine model fit: χ /df <3, CFI and TLI ≥ 0.90, RMSEA <0.08, and SRMR <0.08
(Kline, 2016).
The 4-factor model of the TEIQue-SF, which includes six items for the well-
being factor, six items for the self-control factor, eight items for the emotion-
ality factor, and six items for the sociability factor, did not demonstrate a sat-
isfactory fit to our data based on the CFI and TLI indices (Model 1 in Table
1). This poor model fit can be attributed to certain weak items within the di-
mensions. Specifically, four items (TEIQue_2 and TEIQue_23 from the emo-
tionality factor, and TEIQue_10 and TEIQue_25 from the sociability factor) ex-
hibited low standardized loadings (below 0.40). Consequently, these items
were eliminated, and the CFA was re-run for the Slovenian adapted 4-factor
model, which included six items for the well-being, self-control, and emo-
tionality factors, and four items for the sociability factor. This model (Model
2 in Table 1) showed an adequate fit to the data.
The standardized factor loadings (λ) for the items in the Slovenian ver-
sion of the TEIQue-SF are presented in Table 2. The factor loadings vary
between 0.422 and 0.731, which are sufficiently high and meet the recom-
mended threshold of above 0.40.
The reliability of the Slovenian version of the TEIQue-SF was assessed us-
ing Cronbach’s alpha, which indicates that the overall questionnaire demon-
strates adequate internal consistency (α = 0.88). The individual dimensions
also exhibit sufficient reliability: well-being (α =0.83),self-control (α =0.68),
and emotionality (α = 0.72). However, the dimension of sociability shows a
slightly lower internal consistency with an alpha of 0.55. Due to the favour-
Table 1 Confirmatory Factor Analysis: Fit Indices of k-Tested Models
2
Measurement model χ /df CFI TLI RMSEA ( CI) SRMR
Model . . . . (.–.) .
Model . . . . (.–.) .
128
