Public stigma towards informal caregiving in Germany: a descriptive study

Abstract Objectives This study aims to investigate the public stigma towards informal caregivers of individuals aged 65 years and older in the German population. Method In an Online-Survey 1038 participants (18 years and older and living in Germany) were sampled using a quota-system based on the German micro census data. To assess public stigma towards informal caregiving for individuals aged 65 years and older, three aspects of stigma were assessed: 1) emotional reactions (adapted Emotional Reaction to Mental Illness Scale), 2) behavioral reactions (adapted Social Distance Scale), and 3) cognitive reactions (newly developed List of Cognitions regarding informal caregiving). Exploratory and confirmatory factor analyses were conducted for the development and adaptation of the three instruments, and their composite reliability is provided. Results Emotional reactions in terms of devaluing feelings and feelings of misery were low, while appreciative feelings were neither low nor high. Reported social distance was low. Agreement with cognitions towards informal care in terms of devaluing and accusative cognitions was on average low, but high regarding appreciative cognitions. Conclusion The results show that there is stigma towards informal caregiving for older individuals (65 years and older) in terms of emotional reactions, cognitions and social distance, although the magnitude of this stigma seems to be small. Knowledge regarding the occurrence and magnitude of stigma, as well as the processes that underlie this, is needed to prevent stigmatization and its negative outcomes, or at the very least, to enable the development of support for the affected individuals.


Introduction
In Germany, like in many other industrial countries, a demographic change is occurring, leading to a growing number of individuals in higher age groups (Federal Statistical Office of Germany, 2011). Care at home and by relatives has been reported to be preferred (Hajek, Lehnert, Wegener, Riedel-Heller, & K€ onig, 2018;Heuchert, K€ onig, & Lehnert, 2017), thus, the demand for informal care (private care provided by relatives and friends) is expected to increase. However, informal care can have detrimental consequences for caregivers, such as worse health outcomes (Allen et al., 2017;Zwar, K€ onig, & Hajek, 2018), as well as for care recipients such as reduced life satisfaction (Ouyang, Sun, & Wang, 2019;Zwar, K€ onig, & Hajek, 2019). Although many studies have investigated these consequences, very few studies have looked into the evaluation and treatment of informal caregivers by the population in terms of stigmatization.
Stigma can be defined as a negatively perceived attribution of an individual, which leads to a change of the perception of the stigmatized person (Goffman, 1967). Two main perspectives can be differentiated: the perspective of the stigmatized individual and the perspective of the individual displaying the stigma towards the stigmatized person. From the perspective of the stigmatized, one can assess perceived stigma as it is perceived by the stigmatized person, or self-stigma, which refers to the internalization of the perceived stigma (Werner, Mittelman, Goldstein, & Heinik, 2012). From the perspective of the person displaying this stigma one can assess public stigma which is the stigma displayed towards the stigmatized individuals as it is reported by society (Corrigan, Markowitz, Watson, Rowan, & Kubiak, 2003) and is the focus of this study.
Previous research on informal care has mainly focused on a perceived stigma type, namely courtesy stigma, which refers to the stigma that is transferred from the care recipient to the caregiver, due to their association (Angermeyer, Schulze, & Dietrich, 2003;Goffman, 1967). For example, Abojabel and Werner (2019) found negative cognitions, emotions and social distance, as well as some positive emotions towards the caregivers of individuals with Alzheimer's disease in their qualitative study. Further research indicates women and adult children report higher perceived courtesy stigma than men and spouses, respectively (Kahn, Wishart, Randolph, & Santulli, 2016). Another qualitative study which investigated the internalized courtesy stigma of caregivers found negative emotions as well (Navab, Negarandeh, Peyrovi, & Navab, 2013).
Regarding stigma towards informal caregivers of veterans, findings indicate the occurrence of stigma towards informal caregivers (Phelan et al., 2011) and negative consequences of this stigma for the caregiver's health and social integration (Phelan et al., 2011;. Adding to this, increased critical behavior of caregivers towards care recipients was indicated when perceiving greater stigma (Weisman de Mamani, Weintraub, Maura, Martinez de Andino, & Brown, 2018).
In sum, qualitative and quantitative studies have found evidence of the occurrence of courtesy stigma towards informal caregivers (Abojabel & Werner, 2019;Kahn et al., 2016;Navab et al., 2013). However, these studies investigated the perspective of the caregiver, namely, how the caregiver perceived stigma aimed from the society towards them (and their care recipients). However, perceived stigma, and public stigma that is actually displayed by society, may differ. Only by exploring both perspectives will it be possible to gain a comprehensive knowledge of the existence and magnitude of public stigma. This knowledge is needed for developing interventions to prevent or decrease stigma. Furthermore, these studies almost solely focused on informal care of individuals with Alzheimer's disease (e.g. Abojabel & Werner, 2019;Kahn et al., 2016;Navab et al., 2013), with the exception of the studies from Phelan and colleagues (Phelan et al., 2011;. Due to this, the results cannot necessarily be generalized to all informal caregivers. Thus, there is a gap in the research literature with respect to the occurrence and magnitude of public stigma towards informal caregiving. However, in view of the possible negative consequences of stigmatization of informal caregivers that have already been found (Phelan et al., 2018;Weisman de Mamani et al., 2018), the investigation of stigma towards informal care is highly relevant. Moreover, in addition to these negative consequences, stigma could also discourage individuals from caregiving or lead them to hide their caregiver burden (Phelan et al., 2011). It could also stop caregivers from looking for support (Werner & Heinik, 2008), as has been witnessed in other research fields on stigma (Schnyder, Panczak, Groth, & Schultze-Lutter, 2017). This can endanger the caregivers by worsening their situation. Given the expected increase in demand for informal care, public stigma towards informal caregivers may endanger the capability and availability of informal caregivers.

Conceptual framework
Informal care in this study is defined as performing private care (e.g. relatives, friends or neighbors) for a person aged 65 years or older, including personal care, help with feeding, mobility, or household chores during the last six months at least once per week.
To define stigma, we refer to the definition by Goffman (1967) and the stigma theory from Link and colleagues (Link, Yang, Phelan, & Collins, 2004;Link & Phelan, 2001), which we apply to the caregiving context. Stigma refers to a negatively perceived attribution of an individual (Goffman, 1967). The development of stigma (Link et al., 2004;Link & Phelan, 2001) starts with labeling individuals with a specific characteristic or behavior, and distinguishing them from those without it. The labeled group is seen as separate from others and specific cognitions or stereotypes are associated with the label, which are in turn associated with emotional reactions. In consequence, these individuals are treated differently and discriminated against. Applying this theory to the caregiving context, we assume that individuals are labeled due to providing informal care for older individuals. This label is associated with cognitions and emotional reactions, which in turn lead to discriminating behavioral responses, for example, by distancing themselves from caregivers.
As mentioned previously, various forms of stigma exist. This study focusses on the public stigma as it is demonstrated and perceived by the society. The reason for this focus is that, by definition, stigma is based on social evaluation, and the public stigma influences other forms of stigma, like self-stigma (Corrigan et al., 2003;Corrigan & Watson, 2006;Goffman, 1967). Due to this, we see public stigma as a central stigma concept.
Therefore, the aim of our study is to investigate if and to what extent public stigma towards informal caregiving exists in the German population.

Design and subjects
An Online-Survey was conducted with a computer assisted web interview (CAWI, approximate duration 10 min) among the adult population (18 years and older) living in Germany. Participants were recruited online (e.g. social media networks, opt-in-mails, and co-registration) using a quota-system based on the German micro census data of 2016. Recruitment and data assessment were conducted by the market research institute USUMA (Unabh€ angiger Service f€ ur Umfragen, Methoden und Analysen (engl. Independent service for surveys, methods and analyses)) in cooperation with Lightspeed GMI (Global Market Insite).
Participants were given a vignette describing a typical situation of informal long-term caregiving for an aged care recipient. After reading the vignette they were asked to complete the questionnaire assessing public stigma in regard to the caregiver described in the vignette, and to provide sociodemographic information. All participants gave informed consent. We applied for an ethics vote at the ethics committee of the medical chamber of Hamburg ( € Arztekammer Hamburg), who declared no ethics vote to be necessary.

Instruments
Before the final data collection by USUMA, three pretests were conducted to test the plausibility, comprehensibility, and operability of the questionnaire. Pretest 1 (N ¼ 8) and pretest 2 (N ¼ 3) involved a convenience sample of academics and non-academics, respectively. The pretests were conducted during the final phase of questionnaire development by the research group. Participants were asked to comment on plausibility and comprehensibility of the questionnaire, and the questionnaire was adapted accordingly. Another technical and practical pretest, pretest 3 (N ¼ 66), was conducted by USUMA, with a sample assessed in the same way as the sample for the final data collection (see description above). During this pretest, participants had to answer to the preliminary CAWI questionnaire online.
Operability, comprehensibility, duration and problems with the questionnaire were analyzed based on these results and the CAWI questionnaire was adapted accordingly.

Sociodemographic data
Participants were asked to give information on their age, gender, marital status (married, living together; married, living separately; divorced; widowed; single), highest educational degree (upper secondary school; qualification for applied upper secondary school; polytechnic secondary school; intermediate secondary school; lower secondary school; currently in school training; without school-leaving qualification) and highest vocational degree (university/university of applied science/school of engineering; college/ technical school/master school; vocational school/apprenticeship; without qualification). They were also asked about their employment status (employed, retired, not employed), if they were born in Germany, in which state in Germany they live and how religious (0 ¼ not religious to 100 ¼ very religious) they are.

Experiences with caregiving
Participants were asked if they had experience with providing informal care, and if they had friends or relatives with experience with informal caregiving.

Public stigma
Three questionnaires were implemented to assess public stigma towards informal caregiving for individuals aged 65 years or older in the German population in terms of emotional, behavioral and cognitive reactions. To cross-validate the instruments, the total sample was randomly split in two samples. An exploratory factor analysis was calculated for each instrument with sample 1 (N ¼ 500), and a confirmatory factor analysis was conducted for each instrument with sample 2 (N ¼ 538) (further information on the two samples available on request). Composite reliability was calculated for each factor.
Emotional reactions towards informal caregivers. The first instrument aimed to assess the emotional reaction towards informal care. So far, no research on emotional reactions towards informal caregiving for older people has been conducted. Therefore we used an adapted version of the Emotional Reaction to Mental Illness scale (ERMI) (Angermeyer & Matschinger, 1996). Participants had to rate the emotions on a 5-point Likert Scale (0¼'not applicable at all' to 4¼'totally applicable') in regard to the caregiving situation described in a vignette. However, the ERMI has been developed for research on stigma towards mental illness, thus, the assessed emotions are primarily relevant for this context. Different and possibly additional emotions may be of more relevance in the caregiver context. Due to this, we adapted the instrument to assess all basic emotions by adding further items in accordance with one of the leading theories on emotion, the theory of basic emotions from Ekman (Ekman, 1992;Ekman & Cordaro, 2011). The following emotions were added: sadness, disgust, guilt, shame, embarrassment, happiness, rejoicing, wonder, surprise, envy, pride, relief, contempt, excitement, and schadenfreude. In discussion with two native English speakers we translated rejoicing into admiration. Compassion was added, despite not being defined as a basic emotion, because compassion is of central relevance and often expressed towards stigmatized groups (Katz, 1981). Based on the pretests we excluded some of the emotions: Ecstasy was described as an inappropriate emotion in this context; one item from the ERMI assessing anger and two items assessing fear were found to be redundant and therefore excluded. Due to low variance (<.40) three items were excluded after data assessment (Item 3 amusement; Item 9 disgust; Item 21 schadenfreude).
An exploratory factor analysis (EFA) using principal factor analysis and oblique oblimin rotation was conducted on the final instrument. The Kaiser-Meyer-Olkin measure (KMO¼.862) and the KMO values for all items of the final model were well above .50, and the Bartlett test of sphericity was significant (v 2 (153) ¼ 3048.708; p <.001). Thus, sufficient correlation between the items existed. The Kaiser criterion (eigenvalue > 1) and the scree plot indicated a three factor solution. Two items were excluded because they loaded below .40 (Field, 2013). The repeated EFA without these two items found a three factor solution again (Factor 1: Devaluing feelings (49.99% of common variance explained); Factor 2: Appreciative feelings (43.57% of common variance explained); Factor 3: Feelings of misery (17.74% of common variance explained)). Factor 1 and Factor 2 were negatively correlated (r=-.11), Factor 2 and 3 were positively correlated (r¼.07) and so were Factor 1 and 3 (r¼.05).
Social distance. The second instrument for the assessment of the behavioral reaction was the Social Distance scale (SD), which had originally been developed by Bogardus (Bogardus, 1925(Bogardus, , 1933 and has since been adjusted for use in various contexts (Sowislo et al., 2017;Weaver, 2008). Its results should give indications towards possible consequences of the stigmatization process, namely, the desire for distance from the stigmatized group. We used the German translation of the modified version from Link, Cullen, Frank, and Wozniak (1987). The SD has good validity and reliability (Cronbach's alpha ¼ .75 -.90) and is regularly used to assess stigma (Link et al., 2004). Based on the results of our pretests we adjusted the item formulation to be closer to the original questionnaire by changing the items back to statements, and added two items. Results from the pretest suggested this change in wording to improve the comprehensibility of the items. The 9 items had to be rated on a 5-point Likert scale (0¼'do not agree' À 4¼'totally agree') in regard to the caregiving situation described in a vignette.
An EFA using principal factor analysis was performed after the tests of the assumptions showed that the KMO measure (KMO¼.939) and values were well above .50 and Bartlett's test of sphericity was significant (v 2 (36) ¼ 2916.336; p <.001). The factor analysis revealed a onedimensional structure of the instrument (Factor 1 explained 102.30% of the common variance).
This model was tested with a CFA with bootstrapped standard errors (replications 1000) with results for the model (v 2 (27) Despite RMSEA being above the recommended cut-off of .08 for a reasonably close fit, SRMR is well below the cutoff of .05 or .08 and CFI is above the recommended cutoff of .90 or .95, as well as TLI achieving the cutoff of .95. Thus, the model is accepted as a good fit (coefficients and bootstrapped standard errors are given in Appendix A, Table A2). A composite reliability of .92 was found, indicating excellent reliability.
Based on this result a mean score was built from the 9 items, with higher scores indicating less social distance.
Cognitions regarding informal caregivers. The third instrument was used to assess the cognitive reaction and included a list of possible statements (referred to as cognitions) regarding informal care for older individuals (65 years and older), for example, 'Caregivers provide care because they want to feel needed'. Subjects had to rate these cognitions on a 5-point-Likert scale (0¼'not applicable at all' to 4¼'totally applicable') in regard to the caregiving situation described in a vignette, with higher scores indicating higher agreement with the cognitions. The list was developed for this study based on mostly qualitative research in scientific databases, as well as reports in public media on stereotypes voiced about caregiving for older people in general. In addition, experts in professional care gave advice and the item generation was discussed extensively within the research team. Furthermore, three pretests were conducted to test the plausibility and comprehensibility of the list. Due to low variance ( 40) two items were dropped (Item 7: 'Carers are nauseating', and Item 10: 'Carers provide care because they are too lazy to do real work') after data assessment.

Sample
The survey was completed by 1038 participants. It was quite representative with respect to regional distribution, women were only slightly underrepresented, and men slightly overrepresented. However, in the younger age group of 18 to 34 years of age, men in particular were underrepresented, while in the higher age group of 75 years and older, men were overrepresented and women quite underrepresented. The mean age of participants was 52 years (SD ¼ 16.63, Range: 19-90) and there were marginally less female participants (48.94%). Regarding the marital status 60.40% were married and currently living together with their partner, and 21.19% reported being single. The majority had either an upper secondary school qualification (34.78%) or an intermediate secondary school qualification (31.89%), while only a small number of participants had no school training (0.19%) or were still in school training (0.39%). About half of the sample (52.12%) had an apprenticeship or vocational school qualification, while about a third (27.17%) had a university degree as their highest vocational degree. The majority of the participants were employed (53.18%), with 14.35% being unemployed and 32.47% being retired. About 94% were born in Germany. The religiousness of participants was on average 32.14 (SD ¼ 31.88, Range: 0-100). Previous experience with providing informal care for an individual aged 65 years or older, was reported by 29% of the sample and about a third of the sample indicated they had friends or family who are informal caregivers (34.10%). In general, the variables had very few missing values (between 0.10 and 0.39%, with the largest percentage of missing values found for religiousness: 2.7%) (more detailed information available on request). Details on the sample can be found in Table 1.

Public stigma towards informal care
Results regarding the Emotional Reactions towards Informal Caregivers scale are shown in Figure 1 and  Figure 2 and Table 3 show the results of the Social Distance scale. Individuals reported on average low social distance (M ¼ 3.08; SD¼.76). Highest agreement was given to being comfortable with a caregiver as a neighbor (M ¼ 3.39; SD¼.80). Lowest agreement was given to entering a romantic relationship with the caregiver (M ¼ 2.33; SD ¼ 1.23).
Regarding cognitions towards informal caregivers results are displayed in Figure 3 and Table 4. Descriptive analysis show that individuals reported on average low agreement to accusative cognitions towards informal caregivers (Factor 1) (M ¼ 1.05; SD¼.67). Highest agreement was given to Item 12 'Carers provide care because they want to feel needed' (M ¼ 1.82; SD ¼ 1.22), while lowest agreement was given to Item 6 'Carers provide care because they can't do anything else.' (M¼.59; SD¼.98). Participants also showed on average low agreement with devaluing cognitions (Factor 2) (M ¼ 1.22; SD¼.74). Agreement to these cognitions was very similar, with highest agreement being given to 'Caregivers are often depressed.' (Item 5, M ¼ 1.50; SD ¼ 1.11). Appreciative cognitions were on average rated high (Factor 3) (M ¼ 3.32; SD¼.70). Highest agreement was given to 'Caregivers provide a valuable service to society' (Item 9, M ¼ 3.53; SD¼.81); lowest agreement was given to Item 20 'People work as caregivers for the sake of their relatives.' (M ¼ 3.10; SD¼.99).
Correlations between the three public stigma instruments were calculated. Strong associations could be found between appreciative feelings and social distance (r¼.53) and between appreciative cognitions and social distance (r¼.50) (further results not shown but available on request).

Discussion
The purpose of this study was to investigate the occurrence and the magnitude of public stigma towards informal caregivers of individuals aged 65 years and older in the German population.
As explained by the theory from Link and colleagues (Link et al., 2004;Link & Phelan, 2001), stigma is based on emotional reactions and cognitions or stereotypes, which express themselves in behavioral actions, like social distance. Our first instrument measured these emotional reactions with a three-dimensional instrument. Only a low level of devaluing feelings were reported. Feelings of misery were slightly higher than devaluing feelings, while an even higher degree of appreciative feelings were reported, which was still neither low nor high. This indicates that, while informal care does not seem to be associated strongly with negative emotional reactions, it is not associated with a high degree of positive emotions either.
The results are slightly different regarding cognitions. Here, a clear contrast could be found between deprecating cognitions (accusative and devaluing cognitions), which were ranked lower in agreement, and appreciative cognitions, which had rather high agreement. Social distance was also reported as rather low, indicating a low degree of stigmatizing behavior towards informal caregivers.
Thus, in general, stigma towards informal caregivers, as measured with the emotional, social distance and cognitions measurements that we employed, did occur, but was rather low in the analyzed population.
Former quantitative studies, like Kahn et al. (2016), also found stigma in the specific group of caregivers they analyzed. Further studies reported internalized stigma towards informal care (Navab et al., 2013). Furthermore, qualitative studies like Abojabel and Werner (2019) reported that caregivers perceived various negative and positive emotions and behavior, as well as negative cognitions. However, since many of the former studies used a qualitative study design, their results are only representative for the group investigated. Adding to this, these studies focused only on caregivers of individuals with Alzheimer's disease. Thus, there may also be a difference in stigma in regard to the form of care that is given, with dementia caregiving being more stigmatized. Further research with quantitative study designs are needed to test this.
Moreover, qualitative studies cannot quantify their results and thus give no indication regarding the magnitude of stigma. Therefore, comparing our results with former results regarding the magnitude of stigma is not possible.
Additionally, the difference in the type of stigma that has been analyzed has to be taken into consideration. While former studies investigated the perception of stigma by the caregivers, by asking them about stigmatizing emotions, cognitions or behaviors they had perceived in others, our study investigated the stigma as reported by the population. Internalized as well as perceived public stigma, as has been reported by previous research, may differ to the occurrence of public stigma. Both types of stigma can be biased by the perspective that was investigated. For example, caregivers may be more worried of and more perceptive towards stigma expressed towards them; those displaying stigma may not even be aware that their reactions are perceived as stigmatizing. Therefore, information on both perspectives is necessary to gain a comprehensive understanding of caregiver stigma. Our results extend the findings of former studies on stigma as perceived by the caregivers, by showing that not only do caregivers report to perceive a stigma, a public stigma is expressed by the society as well. This is of high relevance in light of the significance that public stigma has for other forms of stigma (Corrigan et al., 2003;Corrigan & Watson, 2006;Goffman, 1967).
A possible explanation of our results in the context of former findings of perceived stigma may be that informal caregivers would like to be shown more appreciation for the caregiving work they perform, which they perceive to be invisible and unacknowledged (e.g. Russell, 2001Russell, , 2007. In regard to findings on perceived stigma, these feelings of not being appreciated may lead to a more critical perspective on society's reactions and foster the perception of informal caregivers of a stigma of caregiving, as has been reported in former research (e.g. Abojabel & Werner, 2019;Kahn et al., 2016). In regard to public stigma, our results suggest that the stigma may be expressed less in terms of a negative reaction towards informal caregivers and instead as a lack of empathy. This is indicated by our findings of only a moderate magnitude of appreciative emotional reactions. Still, we found a higher degree of positive cognitive reactions and low social distance. Low social distance is not an expression of appreciation though and it is possible that the more positive cognitive reactions are not expressed towards informal caregivers. Thus, the society seems to express a lack of appreciation which may be perceived as indifference and foster the perception of stigma by informal caregivers. Further research is recommended to investigate this in more detail.
Another aspect that should be taken into consideration is that the perceived stigmatization of informal care recipients has been reported to be higher than the perceived stigmatization of the caregiver (Werner & Heinik, 2008). However, although stigma towards the caregivers might be lower, it can still occur. Considering the consequences of stigma that have been shown by former research, like poorer mental health and increased critical behavior towards the care recipients (Phelan et al., 2018;Weisman de Mamani et al., 2018), the existence of even low stigma should not be ignored. Thus, future research is needed to form a better understanding of stigma towards informal caregiving.

Limitations, strengths and future research recommendations
The sample was drawn as an online sample, which restricts the representativeness of our sample, due to availability Note. Mean and standard deviation are given for continuous variables (age, religiousness), frequency and percent are given for categorical variables (gender, education, employment status, marital status, Land of Birth, experience with providing informal care onself, friends or family with experience with providing informal care). and capability of internet access (online affinity) likely influencing the sample selection. However, in online-surveys a social de-contextualization effect has been found (Taddicken, 2009), with individuals answering less frequently according to social norms. Thus, with a sensitive topic like stigma, we might have the advantage of reducing the social desirability bias linked to the assessment of sensitive topics, by using an online survey.
Our study uses instruments that have been adapted and developed for the assessment of stigma of informal caregiving and require further validation. Still, it should be noted that the SD Scale from Bogardus (1925Bogardus ( , 1933) is a valid and reliable instrument, which has been widely implemented in different research contexts. As the results of the CFA for the modified SD scale show, the one-dimensional structure of the original instrument remained stable and the model indices indicated a good fit. Composite alpha was excellent, which suggests high reliability.
We also used the ERMI (Angermeyer & Matschinger, 1996), which is a well-validated instrument regarding stigma of mental illness. We modified it to investigate the emotions relevant for informal-care related stigma and our model was the only acceptable fit. Since further modifications did not improve the model, we assume the assessment of further items might be necessary, especially for Note. Mean and standard deviation are given for the subscales of the Emotional Reactions towards Informal Caregivers scale (devaluing feelings, appreciative feelings, and feelings of misery) and each item of the subscales; participants were asked to rate the emotions of the instrument after reading about a typical caregiving situation (Range 0¼'not applicable at all' to 4¼'totally applicable'), with higher scores (of item and subscales) indicating higher agreement with the emotions.  Factor 3 which consists of only three items. This might also improve the composite reliability of Factor 3. Thus, we recommend to include especially items relating to this factor to improve the model in future research. Lastly, we could not find any instrument assessing specific cognitive reactions towards informal caregiving. Therefore, we had to develop a new instrument. The CFA showed the model was a reasonably close fit, although, the composite reliability for two factors was only moderate. As with the Emotional Reactions instrument, a possible explanation might be the low number of items assessing these two factors. We recommend the addition of more items relating to these two factors of devaluing and appreciative cognitions to improve the instrument in future research. In general, further research on the psychometric properties with a random sample is recommended for all our instruments.
Nevertheless, this study is the first in this new research field to investigate this topic. Only a few studies have been performed in this context yet. Therefore, instruments assessing the public stigma towards informal caregivers specifically have not been available and the modification and development of our instruments was necessary to enable this research. Due to this, our study's results can give a first indication on the occurrence and magnitude of informal care-related stigma in Germany.

Conclusion
Stigma towards informal caregivers looking after older care recipients is a relatively new field of research. So far, only a small number of studies have investigated its occurrence and all have focused on stigma as perceived by the caregiver. However, to gain a comprehensive understanding of this stigma, not only the perceived stigma but also the occurrence and the magnitude of public stigma as shown by the society has to be analyzed. This study investigated public stigma towards informal caregivers, by assessing emotional,  cognitive and behavioral responses to informal caregiving. Thereby, this study gives first evidence regarding the occurrence and magnitude of this type of stigma. The results show that stigma in terms of emotional reactions, cognitions and social distance occurs, although the magnitude of stigma towards informal caregivers seems to be low. In light of the growing importance and demand for informal care, especially in the German care system, these results are of significant importance. Despite the lack of research in this new field, there is evidence indicating detrimental outcomes of stigmatizing for the caregiver (Phelan et al., 2011;, as well as for the care recipient (Weisman de Mamani et al., 2018). Thus, even though the magnitude of stigma we found is low, further research is needed to gain a better understanding of the public stigma towards informal caregiving for older individuals, and the relevant aspects of influence. Knowledge regarding the occurrence and magnitude of stigma, as well as the underlying processes, will be able to help to prevent stigma and its negative outcomes, or at the very least enable the development of support for the affected individuals.  Note. b ¼ standardized factor loadings; unstandardized factor loadings ¼ B; SE b ¼ bootstrapped standard error (1000 repetitions). Appendix A