Design
The data for the validation were obtained from the study “Benefits of Being a Caregiver.” Between October 2019 and March 2020, a total of 50 care assessors from the Medical Service of the Bavarian Health Insurance (MD) distributed 5,000 self-report questionnaires to statutorily insured informal CGs. These CGs applied for an initial grade or an increase in their CRs’ care level at the MD. Of the questionnaires we distributed, 1,082 (21.64%) were returned. By returning the completed questionnaire, the participants gave their consent to the anonymized use of the information they provided.
For the present study, approval was obtained from the ethics committee at the Friedrich-Alexander Universität Erlangen-Nürnberg (No.: 220_20 B).
Sample
After excluding 121 cases because the CRs were younger than 65 years, the current sample for this validation included 961 cases. All CRs were living at home, and 52.8% were living with their CG. The mean age of the CGs was 62.10 years (SD = 12.6), 75.7% were female, 30.5% were spouses, 59.5% were caregiving children/-in-law or other CGs (e.g. aunts, uncles, nieces, nephews), and 47.8% were employed. The mean age of the CRs was 82.12 years (SD = 7.0), and 66.9% were female. A total of 37.9% of the CRs were receiving care because of dementia. Sample characteristics are given in Table 3.
Instruments
Benefits of Being a Caregiver Scale (BBCS) – scale development and description
The seven-step development of the BBCS included several methods: literature reviews, content analysis, surveys, and focus groups. The development was co-created by different experts and affected persons (researchers, care advisors, and caregivers). The development process is described in detail in Table 1. The final BBCS measures the benefits of being a CG at home with a 15-item self-assessment scale. The questionnaire was answered by the participants on a 5-point scale (4=strongly agree, 3=agree, 2=neutral, 1=disagree, 0=strongly disagree). The score ranges from 0 to 60 points. Higher scores indicate greater CG benefits.
Further measurements
CGs’ subjective burden was measured with the 10-item short version of the Burden Scale for Family Caregivers (BSFC-s) [15]. Items are rated on a 4-point scale ranging from 0 (not true) to 3 (exactly true). Higher scores on the self-assessment scale indicate greater CG burden.
Positive aspects of caregiving were measured with the Positive Aspects of Caregiving (PAC) scale. The scale comprises 9 items that are rated on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) [3]. Higher total sum values indicate that the rater experiences greater gains from the care situation.
CGs’ general coping behavior was assessed with the COPE 6 questionnaire, which was derived from the Brief COPE questionnaire [16]. The COPE 6 contains two items from each of the three scales problem-focused, emotion-focused, and avoidant coping. The items are evaluated on a 5-point scale ranging from 0 (strongly disagree) to 4 (strongly agree).
The relationship quality between the CG and CR was assessed currently and before the need for care with the following questions: "How would you rate the quality of the relationship between you and the person you support or care for?"/ "How would you rate the quality of the relationship between you and the person you support or care for before they needed your help or support?” The answer options were represented with a three-level pictorial response format, which characterized the quality of the relationship as "positive," "neutral," or "negative." Both items were subsequently dichotomized. Based on the assumption that socially desirable response behavior leads to negative relationships being classified as neutral [17], the categories "neutral" and "negative" were combined.
The three aspects of informal care time—ADLs (Activities of Daily Living), IADLs (Instrumental Activities of Daily Living), and supervision—were each evaluated with one item according to the Resource Utilisation in Dementia (RUD) questionnaire [18]. For each of the three aspects, the average daily number of hours spent giving care was also asked about. ADLs include activities such as dressing or personal hygiene. IADLs refer to activities such as preparing meals or taking medication. In addition, participants were asked whether and, if so, how many hours were needed for "supervision” (e.g. to avoid dangerous situations).
Other measures
Sociodemographic and background characteristics were also assessed. These included, for example, CGs´ and CRs´ age and gender, CGs´ employment and educational attainment, the CG’s relationship to the CR (categorized into spouses and non-spouses), living situation, duration of care in months, reason for care, and CRs´ care level. All instruments were administered in German.
Statistical analyses
Description
We calculated the mean, median, standard deviation, and distribution of the BBCS scores.
Reliability and item analysis
We calculated Cronbach’s alpha as a measure of internal consistency. Cronbach’s alpha was calculated for the overall score. Bortz and Döring [19] recommended an alpha of 0.80 or higher as an indicator that a scale was well-designed. After the item analysis, the difficulty index and discriminatory power were calculated at the item level. While Bortz and Döring [19] recommended a corridor from 0.20 to 0.80 for the difficulty index, the discriminatory power was calculated as a deleted item-total correlation. According to Bortz and Döring [19], a discriminatory power of 0.30 to 0.50 can be classified as moderate and a power of > 0.50 as high.
Factor analysis
To examine the underlying structure of the BBCS items, we performed an exploratory factor analysis. A scree plot depicts the distribution of the eigenvalues of the individual factors. If more than one factor was obtained with an eigenvalue greater than 1.0 (Kaiser criterion), an orthogonal rotation was used to achieve a simple structure of the grouping of the items. We defined a factor loading ≥ 0.50 as the criterion for assigning a variable to a factor.
Validity
The following hypotheses (H) were tested with regard to convergent (H1 – H3) and discriminant validity (H4):
H1: Because the PACS and BBCS measure similar constructs, the two scales were expected to be strongly positively correlated [3].
H2 : Some authors see a positive evaluation of benefits of being a CG as a meaningful coping resource in and of itself [8], whereas others consider it to be a mediating variable [20]. However, all studies have demonstrated at least a small correlation between these two constructs. Therefore, CGs´ general coping (COPE 6) was expected to be positively correlated with the BBCS.
H3: Different studies have shown a small positive association between a positive relationship between the CG and CR and more positive aspects of care in daily life (e.g. [21, 22]). Therefore, a positive relationship quality was expected to be positively correlated with the BBCS.
H4: Initial results have indicated no association between subjective burden and the positive aspects of caregiving [12]. Thus, no correlation was expected between the BSFC-s and the BBCS.
To test H1, H2, and H4, the correlations between the BBCS score and the metric variables were computed as Pearson correlation coefficients (r). According to Döring and Bortz [23], correlations greater than 0.50 are considered strong, those between 0.30 and 0.50 are moderate, and those between 0.10 and 0.30 are weak. Correlations of less than .10 indicate that there is no association. In order to test H3, eta was calculated to identify the association between the BBCS score and the dichotomized relationship quality variable. According to Döring and Bortz , Eta² greater than 0.14 are considered strong, Eta² between 0.6 and 0.14 are moderate, and Eta² between 0.01 and 0.05 are weak.
IBM SPSS version 28 for Windows was used for all statistical analyses. The cross-sectional baseline data were included in the analyses. A probability of error (alpha level) of 5% was set, below which statistical significance was indicated.