This section discusses the validity and reliability of the T1DM Self-Management Scale for Parents. The scale was presented to five experts to test content validity. Both I-CVI and S-CVI values should be greater than 0.80 to ensure that there is an agreement between expert opinions [23, 24]. In this study, both item-based and scale-based content validity indices were greater than 0.80, which indicated that the scale assessed the construct sufficiently, and content validity was ensured. Content validity analysis showed that the items in the scale were sufficient and appropriate to evaluate the self-management of children with T1DM from parents' perspective [23, 24].

Following the content validity analysis, the construct validity of the scale was first evaluated with the EFA. It was examined whether the sample size was sufficient for the EFA and whether there was a moderate correlation between the items. The adequacy of the sample was evaluated with the KMO value, and the Bartlett test was performed to see if there was a moderate correlation within the dataset. According to the literature, the Bartlett test value should be statistically significant and the KMO value should be at least 0.60 in order to perform factor analysis [20, 21]. In this study, the Bartlett test result was found to be significant and the KMO value was found to be greater than 0.60. These results showed that the sample was adequate, and the correlation matrix was suitable for factor analysis.

Three subscales with eigenvalues greater than 1 were determined in the EFA. Thus, it was decided that the scale would consist of three subscales. The three-factor scale explained 61.427% of the total variance. The literature suggests that the variance explained in new multi-factor scales should be greater than 50% [20, 21]. It is highlighted that higher total variances indicate stronger construct validity [25, 26]. It was determined that the total variance obtained in this study was greater than 50% and the scale had a total variance explained above the recommended level. These results indicate that the scale has construct validity. The EFA revealed that the factor loadings of the items under three subscales ranged from 0.574 to 0.859. The literature suggests that items with a factor loading above 0.30 should be included in scales. In this study, the factor loadings of the items were greater than 0.30.

These results showed that the scale has a strong factor structure [20, 21]. The analysis showed that the items in the scale can reveal the self-management levels of parents of children with T1DM and sufficiently measure the level of self-management and self-efficacy. The EFA revealed that the scale comprises three sub-dimensions: blood sugar management, health monitoring, and exercise and cooperation. These dimensions are critical areas in diabetes self-management, and the scale has been shown to effectively evaluate and measure them from the parent's perspective. Developing a scale that successfully determines the self-management levels of children with T1DM according to these sub-dimensions, especially from the parents' viewpoint, is expected to make a valuable contribution to the literature.

The CFA was performed to confirm the relationship between the items and the subscales and among the subscales obtained as a result of the EFA. The CFA revealed that the ratio of Chi-square value to degrees of freedom was less than five. Factor loadings in all sub-dimensions were found to be greater than 0.30; fit indexes (GFI, CFI, NFI and IFI) were greater than 0.90, and the RMSEA was less than 0.080. A strong and significant relationship was found between the scale and its subscales. In the literature, model fit indices > 0.90, X2/DF value less than five, and an RMSA value < 0.08 are accepted as good fit indicators [27, 28]. The CFA results in this study are consistent with the values stated in the literature. The results revealed that the data were compatible with the model; the structure determined by the EFA was confirmed; the subscales were compatible with the scale, and the items were adequately related to the subscales. The CFA showed that the items could adequately measure what the subscales intend to measure and that they could successfully measure self-management. The CFA supported the conclusion that the sub-dimensions identified in the EFA blood sugar regulation, health monitoring, exercise, and cooperation were compatible with each other and that the scale was sufficient to measure the intended concept. The CFA demonstrated that the items in the scale could adequately assess children's T1DM self-management from a parent's perspective and that the items and subscales were interrelated, confirming the scale's validity and reliability.

The convergent and discriminant validity of the scale was also examined. The CFA showed that the CR value was greater than 0.70, the AVE value was above 0.50, and CR > AVE in all subscales. These results revealed that the scale had convergent validity. The CFA further revealed that MSV < AVE, ASV < AVE; the square root of AVE was greater than the correlation between subscales, and the scale had discriminant validity [27, 28]. These two results show that the items in the scale are highly correlated with the subscales they are under and with the other items under the same factor and that the items have a low or no relationship with the other subscales and items. This result revealed that the items only measure the construct they are expected to measure without confusing them with other concepts, and that the scale has a strong factor structure [27, 28]. Convergent and discriminant validity analyses confirmed that the scale could successfully identify children with adequate and inadequate self-management from the parents' perspective. This indicates that the scale is a reliable and valid tool for assessing various aspects of self-management in children with T1DM.

After the construct validity of the scale was ensured, reliability analyses were performed. First, Cronbach's alpha was evaluated for internal consistency. The Cronbach's alpha coefficient indicates whether items measure similar characteristics. This value is an indicator of homogeneity in a scale. In newly developed scales, Cronbach's alpha value is recommended to be greater than 0.80 for the whole scale and 0.70 for the subscales [22, 26, 29, 30]. It is stated that the Cronbach's alpha values of scales are relatively high, and therefore, McDonald’s Omega coefficients should also be calculated. In this study, the Cronbach's alpha and McDonald's Omega coefficients of the scale were found to be above 0.80 for the whole scale and above 0.70 for the subscales, indicating that the scale has a very high reliability. The literature emphasizes that high Cronbach’s alpha and McDonald’s Omega coefficients indicate strong internal consistency, meaning that the items are compatible with each other and measure only the construct that is intended to be measured [21, 31,32,33,34]. This result shows that the scale is highly reliable. This also shows that all scale items are consistent with each other and measure the same conceptual structure.

In this study, item-total score correlation, item-subscale score correlation, and the correlation between the total scale score and the subscale scores were found to be positive, statistically significant and above 0.30. These results show that all the items of the scale have a sufficient level of correlation with the total score of their own subscales, the item reliability of the subscales is high, and the scale has a reliable and high internal consistency [21, 31,32,33,34].

In the study, split-half analysis was also performed. The literature suggests that in this analysis, the correlation between the two halves should be at least 0.70; the Cronbach alpha values of both halves should be greater than 0.70, and the Spearman-Brown and Guttman half coefficients should be greater than 0.80 [22, 26, 29, 30]. The results of the split-half analysis were found to be above the values recommended in the literature. These results showed that the scale has a high level of reliability.

Item-total score analysis shows whether the items in the scale measure the construct that is intended to be measured [21, 24]. It is recommended that the item-total score correlation should be at least 0.30. In this study, the item-total score correlations exceeded the values stated in the literature, demonstrating that the items were closely related to the overall scale. This indicates that the items were pertinent to the construct being measured, showed a homogeneous distribution, and exhibited high internal consistency. Measurement reliability is crucial in assessing changes in self-management over time in children with T1DM. The scale demonstrated the ability to detect similar results across different conditions, particularly in children whose self-management scores remained stable. Consequently, the scale was shown to produce both accurate and consistent measurements.

In order to obtain accurate results in scales, there should be no response bias [21, 31,32,33,34]. The response bias analysis in this study showed that all the participants filled in the scale according to their own opinions and there was no bias that would affect the scale results [21, 31,32,33,34].

The scale, which was found to provide accurate and consistent measurements, can be used to evaluate children's self-management. According to the literature, self-management in children with T1DM is crucial for preventing acute and chronic complications. Effective T1DM management also significantly improves children's quality of life. Children with T1DM should adhere to a self-management plan that includes daily blood glucose monitoring, insulin injections, physical activity, and adequate and balanced nutrition to maintain blood glucose levels within the reference range and minimize the risk of complications [35]. Failure to maintain good metabolic control can result in acute complications, chronic complications, and even death [35,36,37].

Accurate and consistent determination of T1DM self-management is crucial for identifying the needs of children and preventing future complications. Using valid and reliable measurement tools is essential for this purpose. This study is significant because it introduces a scale that effectively assesses self-management in children with diabetes, contributing valuable insights to the literature.

There are several limitations of the study. The first limitation is the use of convenience sampling. However, it is thought that the sufficient sample size, the follow-up of patients from different regions of Turkey, the university hospital where the study was conducted, and the homogeneous distribution of the participants can reduce the problems related to generalizability. The second limitation is that the data were collected based on self-report. The results of the study are limited to the information provided by the parents. Finally, the study results are based solely on cross-sectional survey data obtained at a single point in time. It is recommended to confirm the results with longitudinal follow-up studies.