Self-perceived strengths among people who are homeless.

This study examined self-perceived strengths among 116 people who were homeless. Those who had experienced a longer period of current homelessness tended to report fewer personal strengths (r = -0.23). Nonetheless, in spite of their marginalized position in society, the vast majority of participants (114 out of 116) perceived personal strengths. A prior diagnosis with mental illness was not associated with the number of strengths reported, but self-perception of strengths was associated with altruistic orientation. The Values in Action (VIA) taxonomy of character strengths captured many of the responses generated by this population. The most frequently mentioned character categories included social intelligence, kindness, persistence, authenticity and humour. The most frequently mentioned other strengths included personal skills (e.g. music, sports), job skills, intelligence and education. The results have relevance for efforts to build self-perceptions that facilitate escape from homelessness.

Introduction

Many studies have considered the pathologies of people who are homeless. In particular, much research has examined factors such as mental illness, substance use and physical illness (e.g. Burns, Lehman, Milby, Wallace, & Schumacher, 2010; Greenberg & Rosenheck, 2010; Yeater, Austin, Green, & Smith, 2010). These studies are useful because pathologies can impede progress towards well-being. Knowledge gained from such research may guide efforts to ameliorate the pathologies and their effects.

Far fewer studies, however, have examined strengths of people who are homeless. This dearth of research on the topic may exist because of a natural inclination to focus on this population's social and psychological difficulties. Also the study of strengths may seem less relevant than the obvious urgent needs of many who are homeless. Whatever the reason, the scant research attention on strengths among the homeless necessarily limits our understanding of this population and of how we might better facilitate their well-being.

In other populations, strength recognition and use seems to produce benefits. In one study, when study volunteers were asked to use their character strengths in a new way each day for a week, they were on average happier and less depressed than a control group one month and even six months later (Seligman, Steen, Park, & Peterson, 2005). In another study, people who reported higher strength use also reported more progress towards their goals (Linley, Nielsen, Gillett, & Biswas-Diener, 2010). Furthermore, a variety of interventions that identify and develop strengths also increase well-being (Quinlan, Swain, & Vella-Brodrick, 2011).

The benefits of strength recognition and use may generalize to people who are homeless if the need for housing and other basic needs are also met. As a first step, however, it is important to gain some understanding of readiness to discuss strengths and of the nature of strength perceptions among people who are homeless.

Several benefits may flow from a greater understanding of strength perceptions among people who are homeless. First, the knowledge gained may help support providers know more about the types of strengths to affirm among this population. Affirming existing character strengths can help build confidence to make other personal changes (Biswas-Diener, 2010). Also, building up self-perceived strengths can enhance well-being (Proctor, Maltby, & Linley, 2011; Seligman et al., 2005).

Second, a strength orientation may even widen pathways out of homelessness. In a prior analysis, people who were currently or previously homeless were asked about factors that would facilitate or had facilitated escape from homelessness (Patterson & Tweed, 2009). Some of the factors facilitating escape were obvious, including receipt of housing, medical care and food. However, those who had already escaped homelessness reported surprisingly often that strength recognition (i.e. realizing that they had something to offer or realizing their potential) contributed to their escape from homelessness. In fact, this sample more strongly affirmed the value of realizing their self-worth in facilitating escape from homelessness than they affirmed the value of several other factors including substance abuse treatment, treatment of a mental disorder or social support in facilitating escape. This suggests that support providers working with people who are homeless may want to encourage cognitive realizations of personal strength. Encouraging recognition of strengths may be more effective, however, if we know which strengths tend to be salient among people who are homeless.

Third, a focus on strengths among people who are homeless may correct societal misperceptions that hinder positive life change for people who are homeless. In particular, for many members of society, the salience of pathology among people who are homeless may often exclude consideration of their strengths (Cozzarelli, Tagler, & Wilkinson, 2002; Cozzarelli, Wilkinson, & Tagler, 2001). The outgroup homogeneity effect is particularly relevant here (Park & Rothbart, 1982). According to research on this effect, people tend to act as if members of any unfamiliar group are very similar to each other (i.e. assuming that the vast majority of people who are homeless are, for example, psychotic or substance dependent or lazy or have any number of other traits which may characterize some people of almost any group). Learning and then familiarizing people with the diverse and even positive traits of outgroup members may help reduce this negative outgroup homogeneity effect and thereby ease societal integration processes such as finding housing or employment.

Several hypotheses are examined here. First, it is reasonable to expect that not all people who are homeless will accept societal stigmatization. Many people resist the norms and definitions given to them by society (Cozzarelli & Karafa, 1998). Furthermore, even people experiencing traumatic events often experience some positive feelings and self-perceptions (Tweed & Tweed, 2011). Thus, we hypothesized that many participants would have ready self-perceptions of strength. The hypothesis is supported by evidence that people who are homeless are more satisfied with aspects of themselves than they are with the social or material aspects of their lives (Biswas-Diener & Diener 2006). In this study, in order to assess the readiness of these perceptions of strength, participants were asked to describe how they differed from other people who are homeless, and also subsequently asked to describe their strengths. The presence of strength-oriented responses was coded for both questions.

Second, we hypothesized that life circumstances, such as homelessness, that marginalize one from society, would diminish the number of strengths reported. Ongoing marginalization may create difficulty in recognizing one's potential (Stillman et al., 2009), Thus, we expected more experience with homelessness to be associated with diminished reports of personal strength.

Third, self-perceptions of strength may be related to other psychological variables. A prior diagnosis of mental illness could conceivably produce an additional sense of stigma that would suppress reports of strength. Thus, we examined the relation between the number of strengths reported and reports of a prior diagnosis with mental illness. We also examined self-perceptions of altruism. In spite of all the difficulties of being homeless, some people who are homeless report acts of altruism. Involvement in altruistic activities could enhance self-perceptions of competence and strength. Simultaneously, self-perceptions of strength could motivate participants to express those strengths by helping others whom they perceive to be in need or less endowed with strengths. Thus, we examined the relation between reports of an altruistic orientation and self-perceptions of strength.

Fourth, we hypothesized that the Values in Action (VIA) classification of character strengths (Peterson & Seligman, 2004) would capture many of the strength constructs elicited when participants were directly asked about their strengths. The VIA classification system was developed to include character strengths that were recognizable as well as valued both across cultures and across history. Research by Biswas-Diener (2006) provides evidence that the 24 strengths included in the VIA classification are widely recognizable even among culturally dissimilar groups. This classification has been used effectively in a wide range of cross-cultural research (e.g. Park, Peterson & Seligman, 2006; Ruch et al., 2010). Other strength frameworks have also been developed. The Realise2 (Linley & Stoker, 2012), for example, is a self-report measure that relies on a framework of 60 different strengths. The StrengthsFinder 2.0 (Rath, 2007), another measure, relies on a framework of 34 strengths. These other frameworks have value, possibly especially for organizational development and career contexts, The Realise2, for example, includes organizationally relevant constructs such as change agent, esteem-builder and explainer, and the SterngthsFinder 2.0 includes organizationally relevant constructs such as developer, relator and strategic. The VIA framework was designed with less emphasis on relevance to organizational productivity and instead a focus on character (i.e. values that can be put into action, hence the acronym values in action). Thus, in spite of the value of these other frameworks, the VIA framework seemed most appropriate for this context with people who are homeless. However, because of the VIA framework's focus on character, it excludes some constructs that may be perceived as important strengths among people who are homeless (e.g. a technical job skill). Thus, it was important to listen for other strengths that were perceived as important to people who were homeless. These other strengths could encompass a broad array of constructs as suggested by the high numbers of dimensions in the Realise2 and StrengthsFinder 2.0. Because the possible array of noncharacter strengths constructs is so large, it is easier to define these by negation, and we thus refer to them as other strengths or sometimes non-VIA strengths or noncharacter strengths. Fifth, we expected that particular strengths would have salience for people in this situation. For example, two studies with homeless women with children uncovered persistence in the face of difficulty (Banyard & Graham-Bermann, 1995; Thrasher & Mowbray, 1995). One of these reports also noted that many of the women commented on their self-sufficiency and their care for others such as by listening. The VIA character strength classification categories capture these themes (e.g. persistence, love, kindness). Another study focused on homeless youth (Bender, Thompson, McManus, Lantry, & Flynn, 2007). The focus group protocol for that study began the discussion of strengths by asking about strengths that helped them cope with life on the street. The personal strengths uncovered in that study were coping skills (including using interpersonal skills to overcome problems), motivation, attitudes and spirituality. Again, components of each of these could be captured by the VIA character taxonomy (e.g. social intelligence, persistence, spirituality). Based on these prior studies, we anticipated that many of the self-reported strengths would be captured by the VIA character taxonomy.

This exploration of self-perceived strengths holds potential relevance for people assisting those who are homeless. In particular, those helpers could possibly assist people who are homeless by assisting them in recognizing their own strengths. Patterson and Tweed (2009) suggested that recognition of one's own strengths may facilitate escape from homelessness. Thus, support providers could start by highlighting strengths that are found to be most often salient to people who are homeless.

Method

Participants

One hundred sixteen participants were recruited at a homeless shelter in Vancouver, Canada through the use of flyers posted in the foyer and elevators. A signup sheet was kept at the front desk of the shelter, so that residents who saw the flyers could choose to sign up for an appointment and participate. The facility provided 70 beds for emergency shelter and 120 beds for supported housing. The facility provided the basics of subsistence, counselling, referral services, education and pro-bono legal services. Participants were required to be fluent in English and to meet the criteria for being considered homeless. Homelessness was defined as (a) staying outdoors from evening until morning because they lacked acceptable housing, and/or (b) staying in an emergency shelter because they lacked acceptable housing. Each participant signed a consent form and then was provided with a $25 gift certificate at the start of the interview and was told that they could quit at any time. There were 82 men (average age of 42.3, SD = 9.68) and 34 women (average age of 38.8, SD = 8.85). A prior diagnosis of mental illness was reported by 46% of participants. A significant proportion reported a current or past drinking problem (57%) or drug problem (67%).

Measures

Could you tell me about your strengths?

In order to assess the frequency of the different strength responses among the sample, the participants were asked the following question: ‘Could you tell me about your strengths?’. The responses were transcribed and then coded. The VIA framework of character strengths (Peterson & Seligman, 2004) was used as a coding guide. The VIA character taxonomy receives extensive use among positive psychologists. Those who constructed the VIA framework aimed to bring together character strengths and virtues that met a number of criteria. For example, each construct included in their taxonomy was widely recognized across cultures, was morally valued in itself (not just for the product it produces), did not diminish others (e.g. competitive would imply dominating others, so it was not included) and is stable in people (i.e. trait-like). Constructs such as fairness, modesty, gratitude, teamwork, love of learning and zest are included. Some responses of participants in this study did not fit the VIA character categories, so codes were added for noncharacter strengths including technical job skills (e.g. certified plumber, computer skills), personal skills (e.g. good at soccer, a good performer), likeability and intelligence/education.

As the coding system was untried in this population and because the large number of categories meant that many responses would have low base rates, we anticipated difficulty in achieving reliable coding. Thus, five assistants coded each response and the data were treated as five indicators of the latent constructs. The data could thus be analysed similar to the way self-report data are analysed when five different questionnaire items assess a single construct. A score could be assigned to each participant for each construct based on the mean scoring of all coders. Thus, for example, if all five raters coded a particular respondent's response as ‘humour’, the respondent would be given a score of one for ‘humour’. If only three of five coders rated a response as being ‘humour’, then the average score would be 0.60. In this way, clear category members (e.g. ‘I'm funny’ would clearly fit within the ‘humour’ category) could receive a higher score in the appropriate category than would less clear responses (e.g. ‘I am not afraid to take charge of others’ could be coded as ‘bravery’ by some coders, but ‘leadership’ by others, so the participant could receive a partial score in each category). Some of the codes were never used by some coders and rarely used by others, suggesting that no clear examples of those responses emerged (e.g. the teamwork and gratitude codes were never used. Curiosity, perspective, modesty, prudence and appreciation of beauty were seldom used). Reliability was assessed with Cronbach's alpha. If a low alpha indicated that a particular coder misunderstood a category, that coder's responses were ignored for that construct (as would be done with aberrant items designed to measure a particular construct on a self-report questionnaire). Cronbach's alpha averaged 0.85 for codes that were used by at least 60% of the coders, with the lowest alpha being 0.66 for ‘likeability’. The responses varied in length. The number of words in response to the question was correlated with number of strengths that were coded (r = 0.461, p < 0.001). This relation makes sense because people who perceive more personal strengths will tend to have more to say in response to a question about their strengths.

How are you different from others who tend to be homeless?

Also, in order to assess the readiness of strength self-perceptions, participants were asked a question that did not mention strengths: ‘This is a bit of a strange question, but when you think about the types of people who tend to be homeless, how are you different from others who tend to be homeless?’. Responses were coded into one of seven categories: (1) character strengths as specified in the VIA character taxonomy (Peterson & Seligman, 2004), (2) historical signs of status (e.g. ‘I had great a job as a marketing executive’, ‘I come from a good family’), (3) lacking the noticeable stereotyped signs of homelessness (e.g. ‘I don't live on the street’, ‘my clothes are clean’, ‘I don't smell’, ‘I don't go binning’), (4) not abusing substances (e.g. ‘I don't do drugs’), (5) not having mental illness (e.g. ‘I'm not mentally ill’), (6) taking steps to overcome homelessness (e.g. ‘Everyday I do one thing to get out of my situation’) and (7)not different (e.g. ‘I'm no different’). This question was asked prior to the strengths question, and thereby assessed the readiness to report strengths even when not directly prompted to do so. The procedure for scoring and for assessing reliability was similar to that used for the strength coding. Cronbach's alpha ranged from 0.80 to 0.96 and was 0.83 for character strength, 0.92 for historical signs of status, 0.91 for lacking signs of homelessness, 0.95 for not abusing substances, 0.90 for not having mental illness, 0.80 for taking steps to overcome homelessness and 0.96 for not different.

Extent of experience with homelessness

Participants were asked questions to assess the extent of their experience with homelessness. In particular, they were asked: (a) how many days since they had their own home, (b) how many years of their life they had been homeless for at least part of the year, (c) how many days of the past six months they had been in a homeless shelter and (d) how many days in that same time they had slept outside.

Prior diagnosis with mental illness

Participants were asked whether they had ever been diagnosed with a mental illness.

Altruistic orientation

Participants also responded to seven statements about altruistic orientation (In the last month, I have contacted a friend to make sure he/she is doing ok, given away food, tried to help someone by listening to his/her problem, worked as a volunteer, wished I could be of help to people, wished I could do something good for other people, hugged someone; alpha = 0.71). The items were generated based on prior focus groups in which a different set of participants, who had recently been homeless, reported on acts of altruism (Tweed & Lehman, 2011). A seven-point response scale allowed participants to say how often they had engaged in each activity during the last month.

Results

Hypothesis 1: Many participants would have ready self-perceptions of strength

When participants were directly asked to describe their strengths, all but two participants described strengths that they possessed. This was the case even though participants were not told prior to the interview that they would be asked to generate personal strengths. Also, earlier in the interview, character strength responses were among the most common response types when participants were asked to describe how they differed from other people who are homeless (Table 1). This response pattern suggests that character and other strengths are important components of selfidentity for these respondents.

Table 1.

How are you different from other people who are homeless?

			Simple contrast with character strength response
Perceived difference from others who are homeless	M (95% CI)	SD	F(df)	p-value	η2
Have a character strength	0.211 (0.139–0.283)	0.383
Not different	0.234 (0.162–0.306)	0.381	0.19(1, 108)	0.664	0.002
Do not abuse substances	0.179 (0.114–0.244)	0.346	0.42(1, 108)	0.517	0.004
Do not have obvious signs of homelessness	0.164 (0.105–0.223)	0.314	1.02(1, 108)	0.316	0.009
Have historical signs of status[*]	0.127 (0.070–0.184)	0.301	4.36(1, 108)	0.039	0.039
Taking steps to overcome homelessness[**]	0.075 (0.032–0.118)	0.227	11.56(1, 108)	0.001	0.097
Do not have mental illness[***]	0.047 (0.010–0.084)	0.196	21.07(1, 108)	<0.001	0.163

Notes: If a participant gave a response that clearly fit within a category (e.g. a character strength), then the participant received a score of one for that category. If the response did not fit the category, a score of zero was given for that category. As described in the manuscript, scores between zero and one were possible. The mean score for each variable (e.g. character strength), thus indicates approximately the proportion of participants who gave that response type. For the significance test, a simple contrast was conducted comparing the character strength response scores to the scores for each other category of response.

p < 0.05, **p < 0.01, ***p < 0.001.

Hypothesis 2: Homelessness may diminish the readiness to perceive personal strengths

In order to assess this hypothesis, we calculated the correlation between total number of strengths participants generated when asked to describe their own strengths and four indicators of extent of experience with homelessness. Gender and age were not associated with the presence of a strength response (p > 0.20 in both cases), so all participants were included in the same analysis. The results are shown in Table 2. The length of time since having a home was associated with fewer strengths being reported. The other homelessness variables were not associated with the number of strength responses.

Table 2.

Correlations between number of strengths reported and four indicators of experience with homelessness.

Indicator of experience with homelessness	r	(95% CI)	p-value
Total days of most recent homelessness[*]	−0.229[*]	(−0.049 to −0.394)	0.013
Days in prior 6 months sleeping in homeless shelter	−0.018	(−0.199 to 0.164)	0.846
Days in prior 6 months sleeping outdoors	−0.093	(−0.270 to 0.090)	0.321
Number of years in life in which homeless at least part of year	0.015	(−0.171 to 0.200)	0.877

Note: *p < 0.05.

Hypothesis 3: Self-perceptions of strength may be related to other psychological variables

Two analyses assessed relations between number of strengths, mental illness diagnoses and altruistic orientation. Strength reports were not diminished among participants who reported a prior diagnosis with mental illness (t(113) = 0.879, p = 0.381, r2 = 0.007). Altruistic orientation, however, was associated with number of strengths reported (r = 0.34, p < 0.001).

Hypothesis 4: The VIA classification of character strengths would capture many of the strength constructs elicited when participants were directly asked about their strengths

For each participant, the total score for VIA character strength responses (M = 1.21, SD = 0.924) and total for non-VIA strength responses (M = 0.932, SD = 0.677) was calculated. Participants tended to report more VIA character strengths than other strengths (F(1, 115) = 6.171, p = 0.014, ?2 = 0.051), suggesting that the VIA character taxonomy captures most of the responses to the strengths question among this group.

Hypothesis 5: Particular strengths would have salience for people in this situation

As shown in Table 3, some of the response types occurred more frequently than the others. For example, the participants valued their skills (e.g. music, sports, job skills), kindness and social intelligence. Men and women reported a similar total number of strengths (t(114) = 0.714, p = 0.477, r2 = 0.004). Men were more likely than women to report job skills (t(114) = 2.25, p = 0.026, r2 = 0.042), personal skills such as music and sports (t(114) = 3.69, p < 0.001, r2 = 0.106), and education or intelligence (t(114) = 2.10, p = 0.038, r2 = 0.037). Women were more likely than men to report forgiveness (t(114) = 2.23, p = 0.028, r2 = 0.042) and persistence (t(114) = 2.35, p = 0.021, r2 = 0.046).

Table 3.

What are your strengths?

Perceived strengths	M	(95% CI)	SD
Personal skills (not job specific, e.g. music, sports)	0.301	(0.237–0.365)	0.350
Character: social intelligence (humanity)	0.180	(0.125–0.235)	0.304
Character: kindness (humanity)	0.162	(0.113–0.211)	0.267
Job skills	0.155	(0.104–0.206)	0.282
Intelligence/education	0.115	(0.063–0.167)	0.287
Character: persistence (courage)	0.104	(0.058–0.150)	0.251
Character: authenticity (courage)	0.097	(0.051–0.143)	0.254
Character: humour (transcendence)	0.070	(0.029–0.111)	0.223
Likeability	0.047	(0.020–0.074)	0.147
Character: spirituality (transcendence)	0.042	(0.009–0.075)	0.184
Character: open-mindedness (wisdom)	0.041	(0.007–0.075)	0.186
Character: creativity (wisdom)	0.040	(0.011–0.069)	0.159
Character: fairness (justice)	0.038	(0.011–0.065)	0.150
Character: love of learning (wisdom)	0.029	(0.006–0.052)	0.127
Character: hope (transcendence)	0.028	(0.002–0.054)	0.145
Character: leadership (justice)	0.026	(0.000–0.052)	0.143
Character: bravery (courage)	0.019	(0.006–0.032)	0.074
Character: prudence (temperance)	0.015	(0.005–0.025)	0.054
Character: love (humanity)	0.013	(−0.006 to 0.032)	0.104
Character: zest (courage)	0.012	(−0.005 to 0.029)	0.096
Character: self-regulation (temperance)	0.009	(−0.004 to 0.022)	0.073
Character: appreciation of beauty/excellence (trans.)	0.008	(−0.005 to 0.021)	0.071
Character: forgiveness (temperance)	0.008	(−0.003 to 0.019)	0.059
Character: perspective (wisdom)	0.005	(−0.001 to 0.011)	0.032
Character: modesty (temperance)	0.002	(−0.001 to 0.005)	0.019
Character: curiosity (wisdom)	0.002	(−0.001 to 0.005)	0.019
Character: gratitude (transcendence)	0.000
Character: teamwork (justice)	0.000

Notes: As in Table 1, if a participant gave a response that clearly fit within a category (e.g. job skills), then the participant received a score of one for that category. If the response did not fit the category, a score of zero was given for that category. As described in the manuscript, scores between zero and one were possible. The mean score for each variable (e.g. job skills), thus indicates approximately the proportion of participants who gave that response type. ‘Character’ indicates that the construct is a character strength from the VIA taxonomy. For variables from the VIA taxonomy, the parent virtue is shown in parentheses. For example, the last strength on the list is teamwork, and, in the VIA taxonomy, this is considered to be an expression of justice.

Some examples for each category might best illustrate the nature of the responses.

Personal skills, for example, often included musical skill. One participant said, ‘I play guitar. I close my eyes and I tell the truth. And there ain't nothing that can touch that or argue with it. I've not found anything big and bad enough to yet. It's straight out, and that's where I speak my truth; that's where I lay my emotions down; that's where I sing’. Another said, ‘I've been practicing for twenty some odd years and I'm very good at guitar. I'm not saying I'm the best player, it's not about being Eddie Van Halen, because he's always going to be out there. To me it's more like about the Beatles, painting a picture. You know, and setting a mood. So I'm very good at that’. The nonmusical personal skill responses were quite varied. One participant said, ‘I can cook, I can play hockey … I took kickboxing’. Another mentioned being good at math. One other mentioned being a good swimmer and a good dancer.

Social intelligence was also mentioned more than many other response types. These responses tended to focus on listening skills and communicating skills. Most of the social intelligence responses were quite brief, though, and provided few specific examples. One participant said, ‘I'm pretty empathetic and understanding’. Another said, ‘I think I've got good communication skills. I'm a good listener. I'm a good talker. I express myself well. I'm very aware of how I feel. I'm intuitive about other people's feelings as well. That's why I think I would make a good counselor’. Another emphasized his listening ability when he said, ‘I have to say that I'm really good with people. Like even before my addiction, with jobs I had and stuff. Like, I'm good with people. Like, a good listener’. Another said, ‘I can be very diplomatic when necessary. I'm good at talking to people. I'm good at understanding people’.

Kindness was also mentioned more often than many of the other strengths. One participant said that his strength is ‘actually trying to help other people if I can. Like, I will go out of my way to help someone. If I can do something for someone that needs it, I will. I guess that's just the way I was brought up’. Others simply said they were kind or generous or that they liked to help others.

Job skills were also mentioned by many participants. Some mentioned specific jobs. For example, one respondent said, ‘I was a ski instructor and a tourist guide and a taxi driver and a truck driver, and I did a lot of things’. Another said, ‘Well, I'm good at construction, anything along those lines’. Another said, ‘I can do typing jobs. And I can do computer jobs. And I can do all the things, like the office things. Yeah, like that’. Others simply said they were good at their job.

Intelligence and education were mentioned several times as well. One respondent said, ‘Anything to do with studying data and, I don't know, education. When I went to school, I went to high school, that required very little effort. I went to college, I did all the schooling for roofing and I got both my provincial ticket and my national ticket, my masters ticket. I have a degree in small business, I have a degree in structural engineering. I have my BA. And so I guess I'm good at that stuff’. Many specifically used the term intelligent, simply saying they were intelligent. One said, ‘I'm a genius. I've got a 145 IQ. I don't know. That's my strength is I'm pretty intelligent’. Some seemed to assume others would doubt their intelligence; one participant said, ‘My strength is I'm intelligent and well-educated, I suppose. All of that is documented’.

Persistence was also mentioned. One respondent said, ‘I did landscaping and I'm not a lazy person. I work until the job's done. It doesn't matter whether it takes 2 hours or 20 hours. I always finish my job. That's what I've been taught’. The persistence was sometimes linked with recovery. For example, one respondent described the following strength, ‘my stubbornness, I think is a good thing for me, because I'm stubborn in my addiction, so I'm stubborn not to go back. And I'm stubborn to move forward. So I need … like I want this. And when I put my mind to it, I can do it. It's hard, and it's slow, but I'll get there’. And another said, ‘I've got to keep picking up the ball and keep going, right, if you fall. So that's a strength I have. I have faith. I have hope. I have a good outlook, you know. What am I trying to say? I'm not a quitter. I won't quit on this’. Others gave a more general statement about their persistence such as ‘I like the fact that I … refuse to quit’.

Aristotle argued in Nicomachean Ethics that the sign of possessing a virtue is finding pleasure in expressing that virtue. Similarly, empirical research suggests that exercising one's top strengths brings satisfaction (Proctor, Maltby, & Linley, 2011; Seligman, Steen, Park, & Peterson, 2005). Consistent with this suggestion, several respondents spontaneously reported that strength use brought pleasure. After describing their strength, they would say that they liked using that strength. For example, one said, ‘I try to press that positive attitude out towards people who are feeling down or negative because they've had a bad turn or a bad time or whatever. If I can bring a smile to someone's face, I'm happy’. Another said, ‘I've got a lot of potential of doing things with my hands. I used to work for [company name], I used to work for restaurants, I used to work for [company name], I built motors … My boss told me, “Here's a map, you have a list, and here's a pen. The tools are right there on the wall. They're all in the shop. You build me that. You figure it out.” There were only two people on it…. It was a nice feeling’. Another said, ‘I'm an excellent restaurant manager and I'm an excellent cook. Yeah. I've been doing it for 25 years, I like it. People like me, I do it well. I get, you know, praise for that, I get paid usually very well for that. Yeah I love my job and that's really important'. Sometimes, the use of the strength benefited others, and this brought pleasure to the participant. For example, one participant said, ‘I have a talent for making people feel better. That makes me feel good to make people feel better or smile'.

In contrast, disuse of a strength can bring displeasure as illustrated by the participant who said, ‘My artistic talents, I'm not so proud of the fact that I haven't been utilizing them as much as I should'. Another said, ‘I think I have a reasonably good understanding and analytical skills and research skills. And I have quite a bit of empathy, and sympathy, for people …. And I understand quite easy high tech from computer skill, knowledge and so on…. Unfortunately, all this understanding is unused, for different reasons. And that's puzzling but also frustrating'.

Some participants expressed concerns that they had lost their strengths. For example, one participant said, ‘I used to be good at sports but the smoking, which I'm trying to quit, has kind of come in the way of that'. Another said that he used to be able to work very hard, but ‘sometimes when I lose weight I get a little muscle deterioration but I just slow down a bit and it takes me a little longer’. Sometimes that concern specifically addressed loss of mental ability. One participant said, ‘I mean I'm a bit banged up but it takes me a little longer. I've got a knock in the head, so it takes me a little longer to put it together now, but it's still there'. Another said, ‘I used to think I was good at math … I don't think that no more'. A third said, ‘I still have a reasonable amount of brain matter. I'm highly intelligent. Maybe it's a little scrambled right now, but very intelligent'. Some participants spontaneously described other barriers to expression of their strengths. One participant said, ‘I'm a musician, I play and sing. I just don't have an instrument right now'.

Others spontaneously mentioned a need to rediscover their strengths. One participant said, ‘That's a tough question because I'm just starting to find who I am again. I was … for anyone that's never done crack cocaine, you can very much lose who you are. And I'm just starting to find who I am in life again'. Another said, ‘What I'm good at? I don't know. I forgot what I was good at. But you know, it's slowly coming back. I'm just, I'm persistent, I'm helpful, I guess I'm outgoing. I can be totally sensitive'.

Others mentioned that they were trying to build their strengths, such as the participant who said, ‘I try to be very reliable, dependable. I'm doing better. I'm not there yet, but I know I can be. I'm reliable. My strengths, I'm making progress here. I know I can do this. It's hard, but, I'm good at finishing what I do, as soon as I set my mind to it'. One participant said, ‘I'm a little lost and uncertain as I go along, but just, I need somebody to guide me. What I need is a mentor, somebody to guide me, someone to straighten me in the same direction, instead of pulling back'. The self-perceived strengths varied, but almost all participants were able to report strengths.

Discussion

Research on problems associated with homelessness (e.g. mental illness, substance abuse, experiences of violence) has value, but strengths may also deserve consideration. This strengths focus among people who are homeless, is consistent with the work of other positive psychologists who study positive traits and experiences even amidst difficult life circumstances (e.g. Biswas-Diener, 2011; Folkman, 2008; Sin & Lyubomirsky, 2009; Tweed, Bhatt, Spindler, Douglas, & Viljoen, 2011; Wong, 2007, 2011). Positive psychology can be relevant even for people who face tragedy (Wong, 2011).

A growing body of research suggests that benefits flow from strength recognition and use. People who report that they use their strengths also tend to report progress towards their goals (Linley et al., 2010) and tend to report higher well-being than do others (Proctor et al., 2011). Furthermore, intervention studies suggest that the effects are causal. In other words, a focus on strength development causes increases in well-being (Quinlan et al., 2011), and that increased well-being can last up to six months even for a brief intervention that involves no therapeutic supervision (Seligman et al., 2005).

Efforts to recognize and develop strengths in people who are homeless could possibly offer similar benefits when combined with provision of housing and other supports. However, little has been reported about strength perceptions among people who are homeless.

This study provides some guidance regarding strength perceptions and readiness to discuss strengths among a sample of people who were homeless. In particular, this study suggests that most people who are homeless can report self-perceptions of strength. When directly asked to describe their strengths, 114 out of 116 participants described a personal strength. When participants were asked to describe how they differed from other people who were homeless, character strengths were among the most common response types. The VIA character taxonomy (Peterson & Seligman, 2004) captured many of the strength perceptions. The top VIA character categories reported were social intelligence, kindness, persistence, authenticity and humour. Some of the most common other strengths (i.e. not character strengths) related to personal skills, job skills and intelligence. Men and women showed some differences in strengths reported. In particular, men more often mentioned skills and intelligence than did women. In contrast, women more often mentioned forgiveness and persistence than did men. Possibly, the men in the study focused their identity on impersonal activity more often than did the women, and, in contrast, the women in the study may have more often found identity in their relationships, and so valued the relationship repairing strength of forgiveness. These are speculations, however, and do not explain why the women more often reported persistence.

Aristotle (Nichomachean Ethics) argued that once any virtue is truly acquired, its use may bring pleasure. Similarly, in this study, several participants closely associated pleasure with use of character strengths. They described their character strength, and then, unprompted, continued to say that it made them happy.

Recognizing and describing one's own strengths can be a powerful experience. Many of the participants became more animated and appeared happier when given a chance to describe their strengths. Some took this as an opportunity to entertain the interviewer. For example, one participant joked, ‘Yeah. I used to be very good with women'. Another told a funny fictional story about two homeless men who travelled the world and hosted a barbecue on the steps of the Vatican. Another respondent described positive comments that a university instructor had made about his writing. Even though the participant owned only a few bags of goods, he had kept that paper and later brought it to show the interviewer. Another participant returned the next week and said that he had felt better all week because of the conversation he had had with the interviewer. Admittedly, not all of the self-perceived strengths mentioned by the participants would be recognized as strengths by others. For example, one participant said, ‘I'm good at dating lots of women'. Another said, ‘There's nothing I've ever tried and failed at. Even as a drug addict, I was a good one’. Another participant mentioned during the interview that he was a very good thief who would never get caught. Another said that the drugs he produced were of higher quality than any available in the region. However, for the majority of strengths mentioned, appropriate expression of these strengths could benefit both the participant and others.

The character strength ranking in this study can be compared to the ranking that Park et al. (2006) reported among Canadian adults. Comparisons must be made with caution because of the many differences between the studies, including the fact that this study used open-ended questioning rather than rating scales. Nonetheless, the differences are interesting. Social intelligence ranked highest among the character strengths with the homeless, but much lower (14th out of 24) among Park et al.'s (2006) sample. Homelessness often requires significant interaction with strangers in unfamiliar and dynamic circumstances and may also provide many opportunities to listen to others’ problems. Thus, this could reinforce a sense of social intelligence both in relation to overcoming social difficulties and also in listening to others. The social intelligence finding supports prior research suggesting that people who are homeless value the interpersonal skills they use to overcome problematic situations (Bender et al., 2007). Also, persistence ranked highly among character strengths with the homeless (3rd), but lower in the typical sample (18th). This finding matches prior reports that people who are homeless value their strength of persistence (Banyard & Graham-Bermann, 1995). Perhaps the context of homelessness highlights the importance of persistence. For many within that context, success will require significant persistence in seeking appropriate supports and facing bureaucratic barriers. In contrast, teamwork and gratitude were never mentioned by the participants, and curiosity ranked low (22nd) among the homeless sample. These ranked higher among the typical sample (15th, 7th and 3rd). Gratitude could be understandably difficult to retain when one is homeless. Similarly, the pressing needs of survival when homeless could press out time for curiosity. Likewise, teamwork seems difficult to muster in this context when projects are very individual (finding food, housing and other supports). These musings about differences between the homeless and typical samples are admittedly speculative, but may suggest hypotheses for future research.

The altruism findings here are interesting, but must be treated with caution. A number of studies suggest that altruistic behaviour can enhance well-being (e.g. Schwartz & Sendor, 1999). Similarly, here, increased reports of strengths were seen among those reporting a more altruistic orientation. However, because of the many stressors faced by people who are homeless, one cannot place expectations of altruism on all members in this group. Thus, this finding should be treated as preliminary and interpreted cautiously.

Similarly, caution is warranted in regard to this whole positive approach with this population. Positive psychology interventions may sometimes cause more harm than good if the recipients are not initially receptive (Lyubomirsky, Dickerhoof, Boehm, & Sheldon, 2011). Furthermore, if a person who is homeless has a simple need for housing, then diverting attention to strengths and other psychological issues may distract from this immediate need. Also, focusing on strengths identification (rather than strength development) can create the impression that strengths are unchangeable (Louis, 2011), and may cause people to give up developing their strengths. This situation would be especially problematic among people who feel that they have few strengths, which could be the case when homelessness suppresses people's perception of strength. Also, inappropriate strength interventions could lead to contextually inappropriate use and even overuse of character strengths (Biswas-Diener, Kashdan, & Minhas, 2011).

Also, practitioners must be sensitive to cultural issues. Some cultures may tend to cultivate a preference for building strengths (e.g. United States, United Kingdom, Canada), but others a preference for minimizing weaknesses (e.g. France, Japan, China; Hodges & Clifton, 2004). Thus, not all participants will be receptive to strength-focused interventions. Possibly, in applied settings, strength-focused interventions could be optional or at least remain only one of a number of intervention strategies used.

Nonetheless, attention to strengths may help many people who are homeless. Some evidence suggests that strength recognition and use is associated with well-being (Proctor et al., 2011) and may even cause lasting increases in life satisfaction (Seligman et al., 2005; see also Rust, Diessner, and Reade; 2009). Biswas-Diener et al. (2011) go further and suggest that use of strengths is not enough, but rather competent use is necessary. They recommend teaching social sensitivity in strength use. In some contexts, use of strengths can cause social problems. For example, when applying for financial benefits, creativity in completing the application can cause problems for the self and others. Thus, recognition and use of strengths can have value especially when the strengths are used appropriately and competently.

However, these strength-perceptions may be suppressed by recent homelessness. People who reported a longer duration of recent homelessness tended to report fewer strengths. In the context of homelessness, mere recognition of strengths may have value. In a prior study (Patterson & Tweed, 2009), people who had been homeless reported that recognizing their own worth and ability had facilitated their escape from homelessness. That sample more strongly affirmed the value of realizing their self-worth in facilitating escape from homelessness than they affirmed the value of substance abuse treatment, treatment of a mental disorder or social support.

For this reason, support providers may want to focus on affirming strengths that tended to have been most readily recognized by participants in the current sample and encouraging appropriate use of them. This would include social intelligence (e.g. ability to listen and communicate), kindness (especially helping others), persistence (e.g. not giving up), authenticity, humour, personal skills (e.g. music, sports), specific job skills (e.g. construction work, mechanical work, workplace computer skills) and intelligence. The gender differences suggest that men may be more likely than women to appreciate recognition of their skills including music skills and technical job skills. Explicitly recognizing, affirming and encouraging appropriate use of strengths among people who are homeless may provide significant encouragement and could possibly bolster motivation to escape homelessness.

Limitations of this study deserve mention. Participants were asked to describe their strengths, so this could be perceived as a demand. Thus, it could be argued that it is not surprising that participants were able to describe strengths. However, two participants did not describe strengths. The two participants who did not generate strengths answered quite briefly. One sounded at a loss for words and said, ‘Nothing. I don't know. I don't know'. The other said, ‘No, because I don't see any right now'. Their inability to recognize their own strengths suggests that findings with the other participants were not simply the result of an irresistible demand. Another limitation is that the strengths were self-perceived. Nonetheless, some evidence suggests that the strength reports have validity. In particular, the relations between strength reports and duration of homelessness and altruistic orientation provide evidence of convergent validity for the strength reports in this study. Other limitations are that the sample came from only one city and had a self-selection bias. Possibly, shelter residents with greater verbal skill and verbal fluency were especially likely to feel comfortable with, and so to sign up for the interview. If this is the case, then participation may be biased towards those residents best able to articulate their strengths. In a more representative sample, open-ended questioning might not have worked as well, and closed-ended rating scales might work better. Thus, there is a reason for further exploration of self-perceived strengths among people who are homeless. In the future, researchers could examine whether strength perception predicts positive change. They could also examine whether strength interventions facilitate positive change. Also, use of a standardized strengths measure (e.g. Peterson & Park, 2009) could allow comparison with other groups.

Conclusion

The study provides evidence that some self-perceptions of strength tend to be retained by and can be discussed with people who are homeless. Self-perceptions of strength may, however, be threatened by homelessness. People who had experienced a longer duration of recent homelessness tended to report fewer strengths.

Helping people who are homeless to recognize these strengths in themselves may offer significant encouragement and could possibly even facilitate escape from homelessness (Patterson & Tweed, 2009). People who had been diagnosed with a mental illness were no more or less likely to report strengths than were the other participants, so strength discussions may be appropriate for diverse groups. Also, attending to strengths of those who are homeless may help the public and service providers recognize residents of homeless shelters as complex people who like others possess many traits including some positive traits (Buber, 1958).

On a more theoretical note, this study also suggests that the VIA taxonomy of character strengths has relevance to marginalized groups. The participants in most prior studies using the VIA taxonomy have been relatively privileged samples or at least not primarily marginalized samples (e.g. Park & Peterson, 2010; Ruch et al., 2010). Thus, some might question its relevance to people who are homeless. Nonetheless, the VIA taxonomy captured many of the responses the participants made when directly asked to describe their strengths. The VIA constructs may thus have relevance for those working with people who are homeless.