Exploring the nature and variation of the stigma associated with loneliness

The current study uses data from The British Broadcasting Corporation Loneliness Experiment to explore the social stigma of loneliness and how it varies by gender, age and cultural individualism. We examined stigmatizing judgements of people who are lonely (impressions of those who feel lonely and attributions for loneliness), perceived stigma in the community and self-stigma (shame for being lonely and inclination to conceal loneliness), while controlling for participants’ own feelings of loneliness. The scores on most measures fell near the mid-point of the scales, but stigmatizing perceptions depended on the measure of stigmatization that was used and on age, gender and country-level individualism. Multilevel analyses revealed that men had more stigmatizing perceptions, more perceived community stigma, but less self-stigma than women; young people had higher scores than older people on all indicators except for internal versus external attributions and people living in collectivist countries perceived loneliness as more controllable and perceived more stigma in the community than people living in individualistic countries. Finally, young men living in individualistic countries made the most internal (vs. external) attributions for loneliness. We discuss the implications of these findings for understandings of loneliness stigma and interventions to address loneliness.

doesn't know what to do to start a friendship or a relationship", "The person is shy", "The person is physically unattractive", "This person believes other people have their own groups of friends and aren't interested in this person"; "This person believes other people are afraid of making friends"; "The person is always in impersonal situations with too many people"; "There aren't enough opportunities to meet people"; "Other people don't try to make friends"; "The person has an unpleasant personality".

Controllability of loneliness. "
If you think about when you feel lonely, to what extent do you agree or disagree that the feeling of loneliness is caused by something you can change?", "If you think about when you feel lonely, to what extent do you agree or disagree that the feeling of loneliness is caused by something you can control?", "If you think about when other people feel lonely, to what extent do you agree or disagree that the feeling of loneliness is caused by something they can change?", and "If you think about when other people feel lonely, to what extent do you agree or disagree that the feeling of loneliness is caused by something they can control?" Perceived stigma in the community. "Below, you will find statements about the community in which you live. Think about your community and say whether you agree or disagree with each statement". Statements were: "In general, people in the community where I live think that people who are lonely are less worthy than others"; "In general, people in the community where I live respect people who are lonely"; "In general, people in the community where I live think there is something wrong with people who are lonely"; and "In general, people in the community where I live tend to think that being lonely is a sign of weakness". Responses were provided on 7 point scales, with 1 = strongly disagree and 7 = strongly agree.
Shame surrounding loneliness. "Think about a time when you have felt lonely. How much do you agree or disagree with the following statements." Statements were: "When I feel lonely, I feel ashamed about it"; "When I feel lonely, I am too embarrassed to admit that to others"; and "When I feel lonely, I do not talk to others about this". Responses were provided on 7 point scales, from 1 = strongly disagree to 7 = strongly agree.
Concealing loneliness. "You have recently started working at a new workplace. One day during the lunch break, one of your colleagues talks about her cousin, who is lonely. She goes into some detail about her cousin's life. Your colleague then begins to talk more generally about people who are lonely. Your colleagues do not know that you are lonely. If you were to find yourself in this situation, having this conversation with your colleague, would you choose to reveal this fact about yourself or would you, instead, choose to conceal the fact that you are lonely?". Responses were provided on a 7 point scale, from 1 = would definitely reveal to 4 = unsure, to 7 = would definitely not reveal.

Supplementary Analyses
Results if feelings of loneliness are not included in the model. In the manuscript text, we include participants' own feelings of loneliness as a covariate when studying the stigma associated with loneliness. After all, it stands to reason that a person who feels lonely themselves may have a different view of the stigma of loneliness than a person who does not personally feel lonely. Here, we report the results when that covariate is not included, that is, when the model includes only the central predictors (gender, age, cultural individualism) and their interactions.

Impressions of people who feel lonely.
None of the effects reached significance. The effect of country-level individualism, which had reached significance when the covariate is included (p=.002) now drops to non-significance (p=.057).
Causal Attributions for Loneliness. Whereas before there were no main effects, now the main effects of age  =0.10, F(1,34310) = 217.16, p<.001 and gender  =0.15, F(1,34310) =148.98, p<.001 reached significance. As before, there was evidence for an interaction between gender and age, F(1,34310)=29.36, p<.001, and the 3-way interaction between gender, age and country-level individualism just reached significance as well, F(1,34310)=8.08, p=.004. Breakdown of these interactions showed the same patterns as before: younger women were the least likely of all groups to differentiate between internal and external attributions for loneliness, and this was especially true in more individualistic cultures. In highly individualistic cultures, young women indicate no preference for internal (vs external) attributions (M= -0.02, SD= 0.72), this differentiates them from young men,  diff =.11, t(34310) = 10.20, p<.001, and from older women, =.004, t(34310)=12.02, p<.001.
No other effects were significant with p < .01. These main effects show that shame was higher amongst women, younger people, and those in individualistic cultures, relative to men, older people, and those in collectivistic cultures.

Controllability of Loneliness
The interaction between gender and individualism, which reached significance before, now dropped to non-significance, F=1. 34, p=.246. No other effects were significant with p < .01. The inclination to conceal feelings of loneliness was stronger amongst younger people, women, and those in more individualistic cultures, relative to older people, men, and those in more collectivistic cultures. No other effects were significant with p < .01.
Analysis using separate impression categories. In the manuscript we take together all impression items into a single scale. However, the original authors (Lau and Gruen, 1992) differentiate four impression categories: Sociability, Adjustment, Competence and General Evaluation. In this section we offer the analysis separated by the different impression categories.
Sociability. Sociability impressions of people who feel lonely were predicted by one's own feelings of loneliness,  = -0.06 , F(1, 8793)= 43.48, p< .001, so that those who feel more lonely themselves report less negative impressions of the sociability of people who feel lonely compared to those who feel less lonely themselves. Additionally, there was a main effect of age,  = 0.06 , F(1, 8793)= 31.68, p< .001, so that older people reported more negative impressions of the sociability of people who feel lonely compared to younger people. No other effects were significant with p < .01.

Adjustment. Adjustment impressions were affected by main effects of all four
predictors, but no interactions. People who felt more lonely themselves, rated people who feel lonely as less well-adjusted than did those who do not feel lonely themselves,  = 0.04, F(1, 8793)= 16.006, p< .001. Women rated people who feel lonely as less well-adjusted than did men, M diff = -0.13, F(1, 8793)= 33.42, p< .001. Older people rated people who feel lonely as better adjusted than did younger people,  = -0.08 , F(1, 8793)= 62.24, p< .001. Finally, those in more individualistic cultures, rated people who feel lonely as less well-adjusted than did those in more collectivistic cultures,  = 0.04, F(1, 8793)= 32.22, p< .001. No other effects were significant with p < .01.
Competence. Competence impressions were affected by main effects of one's own feelings of loneliness, gender, and country-level individualism. People who felt more lonely themselves, rated people who feel lonely as less competent than did those who do not feel lonely themselves,  = 0.06, F(1, 8909)= 27.56, p< .001. Older people rated people who feel lonely as less competent than did younger people,  = 0.07 , F(1, 8909)= 34.68, p< .001.
No other effects were significant with p < .01.

General Evaluation.
Finally, for general evaluation of people who feel lonely, none of the terms reached significance with p < .01.
Taken together, the results on these different impression categories seem somewhat scattered. For instance, there are considerable differences between the effects that appear for ratings of sociability and ratings of adjustment: the effects of age are reversed and so are the effects of one's own feelings of loneliness. The effect of country-level individualism appeared (in the same direction) both for the Adjustment and Competence impressions, and indeed it is this effect that seems most robust, appearing also when taking together all these items into a single scale.

Tables showing the full regression model for all variables The tables below (A-F) show the full multilevel regression model for each of the central
DVs.