Paradox of Self-Stigma: Psychometric Validation for the Portuguese Population in Poverty

Stigma is the attribution of labels to individuals to diminish or weaken them in society. The internalized stigma affects self-stigma, which influences their self-concept and confidence. The aim of this study was to translate and validate the PaSS-24 (Paradox of Self-Stigma Scale) for the population living in poverty in Portugal. The scale was translated using the back-translation technique, and the questionnaire was answered online by 357 people living in poverty in Portugal. The results of the exploratory factor analysis show excellent factorial validity, maintaining the factors of the original scale, and the results of the internal consistency, convergent and discriminant validity show excellent psychometric properties.

Introduction

The situation of poverty affects well-being and mental health, both due to the social stigma associated with this condition and also due to the stigma present in the services that offer assistance to these populations (Ingliset al., 2022). For those whose salaries only allow subsistence, in addition to financial needs, there are a number of associated problems such as marginalization, stigma and lack of access to other resources. If, in addition to poverty, there is an overlap of other minority identities, these can further enhance stigma and increase their vulnerability, with effects on the various aspects of their lives, such as greater difficulty in finding a job and, consequently, greater difficulty in leaving of poverty (Saatcioglu & Corus, 2014).

According to Wilton (2003), poverty appears as an aggravating factor in this context, as these populations have greater difficulty in meeting basic needs such as food, clothing or hygiene products, but also because this condition causes perceived social stigma. This leads to these populations reducing their social relationships and restricting their social relationships, which ends up making it more difficult to reduce social stigma against them, having harmful effects on their self-esteem.

In this context, stigma, characterized by Goffman (1963) as the attempt to diminish an individual through characterizations and disregard of their role in society by giving them negative labels, has a relevant weight here in these more fragile populations. Stigma takes on several aspects, from social stigma to stigma that is perceived by the individual and self-stigma. Perceived stigma and self-stigma are also associated with greater resistance to requests for support, as well as lower levels of perceived social support (Simonset al., 2017). These experiences associated with stigma can cause more vulnerable populations to internalize this prejudice, which undermines their confidence and self-concept and consequently increases self-stigma (Corrigan & Watson, 2002).

For Oexleet al. (2017), self-stigma, in addition to reducing self-esteem, also generates a feeling of shame and lack of hope that can have an impact on suicidal ideation, and in this sense, it is essential to act preventively. Some investigations and instrument validations have been developed on self-stigma, most of which are associated with mental illness, but research still needs further study on this construct: Brown-Johnsonet al. (2015) developed and validated the Internalized Stigma of Smoking Inventory (ISSI) to assess the stigma associated with smokers and whose results indicate that the stigma associated with smoking and associated with cigarette dependence leads to isolation, particularly when it comes to groups with non-smokers. The “Diabetes Self-stigma Scale” developed by Seo and Song (2021) to assess self-stigma in patients with diabetes and can also be used for intervention to reduce self-stigma in these patients and is composed of four factors: Comparative disability, social reserve, self-evaluation and feeling of apprehension. The Stigma Scale of Depression Scale (SSDS) aims to understand self-stigma in people with depression, with a distinction between the stigma of people around them and internalized stigma and can also be used to evaluate interventions on stigma. This scale was developed by Barneyet al. (2010). The Paradox of Self-stigma (PaSS-24), developed by Golayet al. (2021), evaluates the effect of the “condition”, which in the original scale refers to mental health and which, in the adaptation to Portugal, refers to the impact that the condition of poverty has on the internalization of stigma by the person who has the condition in question. There is also a reduced version to assess self-stigma, Self-stigma Scale-Short (SSS-S), validated for the French population by Golayet al. (2022) with 9 items and 3 factors for patients in a psychiatric context.

The objective of this study is to translate, validate and adapt the Paradox scale of Self-stigma (PaSS-24) by Golayet al. (2021) for the population living in poverty in Portugal. The internalization of prejudice and self-stigma are often present in socially fragile populations, and in this sense, it is essential that there is an instrument to measure this construct validated for the Portuguese population.

Method

Participants

This study sampled populations living in poverty in Portugal aged 18 or over and had 357 participants (58.5% women, 41.2% men and 0.3% other gender) aged between 18 and 70 years old (M = 34.51; DV = 14.59). Regarding ethnicity, the majority of participants (79%) identified themselves as “White/White Portuguese/Of European origin”, followed by “Gypsy/Portuguese Gypsy/Roma” (16%) and “Black/Black Portuguese/afro-descendant/African origin” (3.4%), with 1.7% of participants indicating that they belonged to another ethnic group.

The exclusion factor for obtaining the sample was income, in case the per capita family income is higher than the value of the Social Support Index (IAS) in 2023, with a value of 480.43 euros. A total of 501 questionnaires were collected, 143 of which were subsequently excluded as they did not meet the inclusion criteria. The main reasons for exclusion were: Per capita family income higher than the IAS, age under 18, not residing in Portugal and not having answered most of the questions. Descriptive statistics relating to sociodemographic data are presented in Table I.

Table I. Characteristics of Participants

Characteristics	Levels	Frequency (Percentage)
Gender	Woman	209 (58.5%)
	Man	147 (41.2%)
	Other	1 (0.3%)
Marital status	Single/Not dating	103 (28.9%)
	Single/Dating	75 (21%)
	Married/Cohabitation	140 (39.2%)
	Divorce/Separated	31 (8.7%)
	Widow	8 (2.2%)
Ethnicity	White/Portuguese white/Of European origin	282 (79%)
	Black/Black Portuguese/Afro-descendant/African origin	12 (3.4%)
	Gypsy/Portuguese Gypsy/Roma	57 (16%)
	Other	6 (1.7%)
Children	No	175 (49%)
	Yes	182 (51%)
Number of children	1	34 (9.5%)
	2	84 (23.5%)
	3	47 (13.2%)
	4	17 (4.8%)
Education	Up to 4 years	5 (1.4%)
	Up to 6 years	27 (7.6%)
	Up to 9 years	71 (19.9%)
	Up to 12 years	129 (36.1%)
	Degree	101 (28.3%)
	Master’s degree	20 (5.6%)
	Other	4 (1.1%)
Professional situation	Unemployed	120 (33.6%)
	Employee	61 (17.1%)
	Self-employed worker	17 (4.8%)
	Worker/student	11 (3.1%)
	Student	128 (35.9%)
	Retired	19 (5.3%)
	Other	1 (0.3%)
Household	1	21 (5.9%)
	2	31 (8.7%)
	3	77 (21.6%)
	4	133 (37.3%)
	5	59 (16.5%)
	More than 5 people	36 (10.1%)
Household salary	Up to 480.43 euros	34 (9.5%)
	From 480.43 to 960.86 euros	144 (40.3%)
	From 960.86 to 1441.29 euros	132 (37%)
	From 1441.29 to 1921.72 euros	36 (10.1%)
	More than 1921.72 euros	11 (3.1%)
Age (N = 354)	Average	34.51
	Median	35
	Mode	18
	Standard deviation	14.59
	Minimum	18
	Maximum	70
	Range	52

Materials and Measures

Sociodemographic Questionnaire

This questionnaire will be constructed for this research and the questions asked aim to provide a sociodemographic characterization of the participants, such as age, gender, economic situation, educational qualifications and ethnicity.

Paradox of Self-Stigma (PaSS-24). It is an instrument developed by Golayet al. (2021) which assesses self-stigma. This instrument consists of a total of 24 items subdivided into 3 factors: stereotype endorsement (“People with my condition should be banned from certain jobs”), righteous anger (“Certain people’s attitudes towards my condition appal me”) and Non-disclosure (“To stop myself from getting into trouble, I avoid situations where my condition might be revealed”). The rating is made on a Likert scale from 1 (Strongly disagree) to 5 (Strongly agree). The original scale has excellent psychometric properties with a Chronback alpha of 0.93. The scale adapted for the Portuguese population also has excellent psychometric properties with a Chronback alpha of 0.95.

Everyday Discrimination Scale-Portuguese Version (EDS-PT)

The EDS scale was developed by Williamset al. (1997) and aims to measure the experiences of everyday discrimination that individuals suffer for various reasons. The Portuguese version of the scale (EDS-PT) was validated by Seabraet al. (2024). This scale consists of 8 distributed items and two factors: Unfair treatment (items 1, 2, 6, and 7) and personal rejection (items 3, 4, 5, and 8), rated on a Likert scale from 0 to 5, where 0 corresponds to “Never” and 5 corresponds to “Almost always (almost every DAY)”. The scale has good psychometric properties with a Chronback alpha ranging between 0.83 and 0.95.

The Short Version of the Portuguese Positive and Negative Affect Scale (PANAS-VRP)

The PANAS was developed by Watsonet al. (1988) and aims to measure Positive Affect and Negative Affect. The reduced version of the scale (PANAS-VRP) was translated and validated for the Portuguese population by Galinhaet al. (2014) and consisted of 10 items (5 items for measuring positive affect and 5 for measuring negative affect). Items are rated on a Likert scale from 1 to 5, where 1 = Very little or not at all and 5 = Extremely. The scale has two psychometric properties with a Chronback alpha of 0.86 in Positive Affect and 0.82 in Negative Affect.

Procedure

To collect data, an online questionnaire was created and disseminated through social networks and contact with Associations and Institutions that provide support to populations in situations of poverty and with the particular support of EAPN (European Anti-Poverty Network), a partner entity of the study. The sampling was collected by convenience, and all participants were informed of the study objectives, completing informed consent. The inclusion criteria were as follows: Being 18 or over, residing in Portugal, and living in poverty, which was defined by values lower than the Social Support Index (480.43 euros) per member of the household. The collection took place between October 21, 2023, and December 31, 2023. First, a pre-test was carried out to check if there were any problems with any of the questions or any doubts, followed by the collection itself.

The translation and adaptation of the Paradox of Self-Stigma scale (PaSS-24) was carried out using the translation-back-translation technique, using 3 translators of the original scale into Portuguese and asking 3 people to back-translate the scale items. The original author was previously consulted, giving his favorable opinion on the use of this translation and adaptation method. The technique used to validate the scale was convergent and discriminant validity, using instruments that measure similar constructs and others that measure different constructs. In this sense, the Everyday Discrimination Scale-Portuguese Version (EDS-PT) was used for convergent measurement and the short version of the Portuguese positive and negative affect scale (PANAS-VRP) for divergent measurement.

This study is funded by the Foundation for Science and Technology within the scope of the award of Doctoral Scholarships in a non-academic environment (2023.01027.BDANA), and its partner entity is the Vila Real District Center of EAPN–European Anti-Poverty Network/Portugal (EAPN Portugal). The study was approved by the UBI Ethics Committee (Ref. no. CE-UBI-Pj-2023-051-ID1915). The study complies with the ethical standards defined by the Declaration of Helsinki (General Assembly of the World Medical Association, 2014), particularly with regard to anonymity and confidentiality.

Results

For statistical analysis, IBM SPSS version 29 was used. Descriptive statistics, the sensitivity of the PASS items were calculated using measures of Skewness (Sk) for asymmetry and Kurtosis (Ku) for flatness, reliability through Chronback Alpha and content validity through KMO and Bartlett’s Test. An exploratory factor analysis was also carried out using the main components to verify the most appropriate number of factors, as well as the scree plot and analysis of the component matrix to verify the correlation between items and factors. Finally, convergent and discriminant validity was carried out using the correlation between the means of the global scales.

Sensitivity

To analyze the sensitivity of the items, the asymmetry (sk) and flatness (ku) measures were analyzed, checking that all the items were within the expected sensitivity values, in this case, >3 for asymmetry and >7 for flatness. All the items are below the defined values, which indicates good sensitivity of the items, as can be seen in Table II.

Table II. Measures of Asymmetry (sk) and Flatness (ku) of the Paradox Scale of Self-Stigma (PSS) Items

	PSS1	PSS2	PSS3	PSS4	PSS5	PSS6	PSS7	PSS8	PSS9	PSS10	PSS11	PSS12
Skewness	1.16	−1.04	−0.43	0.46	−0.86	−0.37	1.68	−0.67	−0.29	0.95	−0.60	−0.43
Kurtosis	0.48	−0.05	−0.76	−0.97	−0.49	−0.75	2.49	−0.99	−1.20	−0.20	−1.13	−1.05
	PSS13	PSS14	PSS15	PSS16	PSS17	PSS18	PSS19	PSS20	PSS21	PSS22	PSS23	PSS24
Skewness	1.62	−0.85	−0.40	0.28	−0.71	−0.45	1.17	−0.74	−0.91	0.30	−0.79	−0.04
Kurtosis	2.11	−0.52	−1.10	−1.26	−0.83	−1.04	0.59	−0.87	−0.25	−1.27	−0.65	−0.84

Internal Consistency

To assess the internal consistency of the PASS scale, we used the calculation of Chronback’s alpha, which had a value of 0.95 for the entire scale, which reveals excellent psychometric properties. To verify the factor analysis, we used KMO and Bartlett’s Test with a value of 0.95 for measurement adequacy.

For content validity, exploratory factor analysis was used, taking into account that the instrument is being validated in Portugal and as the original instrument is French, there could be cultural differences that would have significance in the way people answer the questions or even in understanding the questionnaire. In this sense, an analysis of the component matrix was carried out to verify whether the number of factors in this sample coincides with the original scale through extraction by main components, as can be seen in the analysis of the figure of three factors is appropriate, as evidenced in the original scale. The first factor explains 49.93%, the second 13.89% and the third 5.85. The scree plot also confirmed the three factors as the most appropriate choice. As can be seen from the correlation of the items with the factors, as shown in Table III, the results show that the items referring to the factors of the original scale coincide in that they all have a higher correlation with the defined factors. In this case, as in the original scale, the factor “righteous anger” contains items 2, 5, 8, 11, 14, 17, 20, and 23, the factor “Non disclosure” contains items 3, 6, 9, 12, 15, 18, 21, and 24, and the factor “stereotype endorsement” 1, 4, 7, 10, 13, 16, 19 and 22 with correlation values ranging from 0.47 to 0.88.

Table III. Exploratory Factor Analysis by Component Matrix

Items	Factor 2–RJ	Factor 3–RR	Factor 1–AE
	Righteous anger	Non-disclosure	stereotype endorsement
PSS2	0.17	−0.03	0.80
PSS5	0.36	0.04	0.76
PSS8	0.31	0.10	0.82
PSS11	0.32	0.15	0.84
PSS14	0.30	0.02	0.83
PSS17	0.31	0.08	0.88
PSS20	0.30	0.07	0.83
PSS23	0.30	0.05	0.84
PSS3	0.66	0.18	0.43
PSS6	0.71	0.15	0.37
PSS9	0.70	0.26	0.42
PSS12	0.75	0.19	0.47
PSS15	0.69	0.22	0.53
PSS18	0.78	0.20	0.45
PSS21	0.70	0.11	0.33
PSS24	0.78	0.20	0.23
PSS1	0.08	0.71	0.23
PSS4	0.39	0.50	0.05
PSS7	0.00	0.82	−0.01
PSS10	0.26	0.76	0.08
PSS13	0.04	0.85	−0.02
PSS16	0.35	0.51	0.50
PSS19	0.37	0.57	−0.19
PSS22	0.38	0.47	0.39

In this sense, Table IV presents the Portuguese version of the PaSS-24 scale (Paradox of Self-Stigma Scale) resulting from this validation, indicating the items per factor, maintaining the factorial structure and the number of items. The instructions given to the respondents for filling out the questionnaire were as follows: “The purpose of this questionnaire is to assess what you think about your mental health. Please indicate whether you agree or disagree with each suggestion. Answer spontaneously without spending too much time on each question. Some sentences may seem strange, jarring or repetitive, but don’t worry. If some suggestions do not match your condition, they may be relevant to others. There are no right or wrong answers; just answer in the way that best describes your feelings.”

Table IV. Portuguese Version of the PaSS-24 Scale (Paradox of Self-Stigma Scale)

Items		Escala
1	As pessoas com a minha condição são menos úteis à sociedade	AE
2	A restrição dos direitos das pessoas com a minha condição indigna-me	RJ
3	Devido à ignorância das pessoas, não falo com ninguém sobre os problemas associados à minha condição	RR
4	Digo para mim próprio: “Para que é que serve lutar por direitos iguais?	AE
5	Estou mesmo farto das ideias preconcebidas sobre a minha condição	RJ
6	Devido aos preconceitos das pessoas, não falo com ninguém sobre os problemas relacionados com a minha condição	RR
7	Certos empregos deveriam ser proibidos a pessoas com a minha condição	AE
8	A falta de conhecimento do público sobre a minha condição indigna-me	RJ
9	Para não me meter em sarilhos, evito situações em que a minha condição possa ser revelada	RR
10	De que serve esforçar-me quando sou inferior aos outros	AE
11	A falta de informação correta sobre a minha condição é escandalosa	RJ
12	Utilizo estratégias para evitar falar da minha condição	RR
13	Certas atividades devem ser negadas a pessoas com a minha condição	AE
14	Os estereótipos sobre a minha condição irritam-me	RJ
15	Para evitar ser prejudicado, utilizo estratégias para evitar falar sobre a minha condição	RR
16	As pessoas com a minha condição nunca terão uma vida feliz	AE
17	A falta de sensibilização dos media para a minha condição é terrível	RJ
18	Para evitar comentários desagradáveis, utilizo estratégias para evitar falar sobre a minha condição	RR
19	As pessoas com a minha condição devem manter-se reservadas	AE
20	Irrita-me a forma caricatural como a minha condição é mostrada na televisão	RJ
21	Para evitar qualquer preconceito, escolho com quem falar sobre a minha condição	RR
22	Aceitei o facto de que nunca poderei ter uma vida social satisfatória	AE
23	A atitude de certas pessoas em relação à minha condição revolta-me	RJ
24	Não revelo a minha condição a ninguém para evitar ser julgado	RR

Construct Validity: Convergent and Discriminant Validity

The scales used to verify convergent and discriminant validity were the EDS and Positive Affect from the PANAS-VRP, respectively. As can be seen in Table V, the results are significant for the average values of the EDS scale, with a positive correlation of 0.67, showing the convergent validity with this scale and are also significant for the discriminant validity, in this case, the Positive affect, with a negative correlation of −0.37. These results demonstrate what was expected for these variables in that the convergent variable presents a positive correlation with the average PSS values, and on the other hand, the discriminant variable presents a significant but negative correlation in relation to the PSS.

Table V. Convergent and Discriminant Validity (Pearson’s Correlation Results)

	Mean_PSS	Mean_EDS	Positive affect
Mean_PSS	1	0.67*	−0.37*

Discussion

The aim of this study was to translate and validate the Paradox of Self-Stigma scale (PaSS-24) by Golayet al. (2021) for the population living in poverty in Portugal, which aims to assess paradoxical self-stigma. The original scale aims to assess paradoxical self-stigma in people with mental problems, using the word “condition” for this purpose, which in the validation for the Portuguese scale was used as “poverty condition” after the scale’s authors were questioned about this change and with due authorization. The translation was carried out using the retroversion translation technique, which maintains the factorial structure and number of items from the original scale. It has a total of 24 items divided into three factors (AE: stereotype endorsement; RJ: righteous anger; RR: non-disclosure), each of them with 8 items. The present scale has good sensitivity and excellent psychometric properties. An exploratory factor analysis was carried out, as there could be differences in the perception of the items due to cultural reasons, as the original scale was adapted for the French population. The results of this analysis were satisfactory, with the scree plot revealing that the three-factor structure is as indicated and, in this sense, it was maintained. The components matrix showed a significant correlation between the items of the same factors, also confirming the results of the original scale.

Stereotype endorsement (AE) refers to the degree of agreement that individuals have with common stereotypes about their condition, such as “People with my condition are less useful to society.” Righteous anger (RJ) comprises items that describe the level of legitimate anger that people have towards stereotypes, such as “The restricted rights of people with my condition is scandalous.” The factor non-disclosure (RR) contains items that express acceptance of the stereotype and the abandonment of trying to reverse people’s view of it, such as “I don’t reveal my condition to anybody to avoid being judged.” The results of the exploratory factor analysis corroborated the factor matrix of the original scale, maintaining the 24 items and 3 factors.

This study thus aims to contribute to research into the situation of poverty and its relationship with self-stigma and how these problems impact the psychosocial health of populations. As Corrigan and Watson (2002) refer, the internalization of stigma, that is, self-stigma, has direct impacts on mental health, namely self-esteem, self-concept and self-confidence. The condition of poverty appears here as an aggravating factor or an enhancer of this self-stigma due to the stereotyping and marginalization that it can cause and also the inherent difficulty in people leaving this situation (Saatcioglu & Corus, 2014).

The possible limitations of this study are related to the fact that the data collection for the sample of this study was done online, and as it is a population in a situation of poverty, there could be a risk of not being able to reach this population due to possible difficulties of internet access. This risk is taken into account, as a study by the National Institute of Statistics (2023) reveals that 89% of Portuguese households have access to the internet at home and that 85.8% have access to a broadband connection. In this sense, it is clear that, as Portugal is a country with practically widespread Internet access, this possible bias does not occur.

On the other hand, as it is a recent scale, there are still no studies that have used the scale in different scenarios. It is essential that studies on this topic continue and that there is continued use of this scale to check for any gaps in the scale.

Conclusion

The Portuguese version of the PaSS-24 scale (Paradox of Self-Stigma Scale), translated and adapted in this study, presented excellent psychometric properties and a good measure for assessing paradoxical self-stigma for people in situations of poverty. The factorial structure and number of items from the original scale were maintained. It is suggested that this scale be used for future investigations into stigma and self-stigma as a way of having a broader view of the internalization of prejudices and, inherently, the best way to act on them. As this is a recent scale, studies on this subject are encouraged to use it so that any weaknesses can be addressed.