Advertisement

Journal of Well-Being Assessment

, Volume 1, Issue 1–3, pp 57–75 | Cite as

General, Health-Specific, and Housing-Specific Self-Efficacy Scales: Preliminary Reliability and Validity Evidence with Homeless or Vulnerably Housed Adults

  • Sneha Shankar
  • Anita M. Hubley
  • Bruno D. Zumbo
Original Research

Abstract

Self-efficacy (SE) refers to one’s sense of personal competence and is a key element of human agency. Among individuals who are homeless or vulnerably housed, SE has the potential to provide important information about an individual’s ability to seek out and make use of resources and persevere in the face of multiple challenges. SE is understudied as a personal variable in research with homeless samples. Thus, it is important to identify appropriate measures of general and domain-specific SE for this population and evaluate their psychometric properties. Three relevant SE scales are the Generalized Self Efficacy Scale, Perceived Health Competence Scale, and Housing Self-Efficacy Scale. The purposes of this study were to examine the internal structure and score reliability for each of these SE scales, report performance on and intercorrelations among the SE scales, and examine the relationship of demographic variables to SE scale scores, with a sample of adults who were homeless or vulnerably housed. Strict unidimensionality was evaluated using confirmatory factor analysis. Model fit was examined using fit indices and residual polychoric correlation matrices. Essential unidimensionality was evaluated using exploratory factor analysis and the ratio of first to second eigenvalues. Internal consistency estimates of reliability were obtained using ordinal alpha. None of the SE measures were found to be strictly unidimensional but all three measures were found to be essentially unidimensional. This finding supported the use of total scores for each measure. Ordinal alphas ranged from .87 to .93 for the three SE measures and thus were satisfactory. Correlations among the three measures ranged from .36 to .45. Demographic variables showed little relationship to the three SE measures. The study findings provided initial psychometric evidence to support the use of these three SE measures with a homeless or vulnerably housed adult sample.

Keywords

Dimensionality Factor structure Homeless Psychometrics Reliability Self-efficacy 

1 Introduction

Pollard and Davidson (2001) defined well-being as “the state of successful performance throughout the life course integrating physical, cognitive, and socio-emotional functions that results in productive activities deemed significant by one’s cultural community, fulfilling social relationships, and the ability to transcend moderate psychosocial and environmental problems” (p. 10) (see also Bornstein et al. 2003). The eudemonic perspective of psychological well-being often includes the following six dimensions: self-acceptance, personal growth, purpose in life, positive relations with others, environmental mastery, and autonomy (Kashdan et al. 2008; Ryff 1989; Ryff and Keyes 1995; Ryff and Singer 1996). Thus, as Ryff and Keyes (1995) note, psychological well-being focuses on whether individuals’ lives have purpose, whether they are realizing their potential, if they feel in charge of their own lives, and what is the quality of their relationships with others rather than strictly if they are happy and satisfied with their lives (i.e., hedonistic perspective; Diener et al. 1999; Watson et al. 1988).

There has been considerable interest in variables that are related to or predictive of psychological well-being. One of those variables is self-efficacy (SE). SE refers to one’s sense of personal competence and belief that he/she can engage in a course of action to bring about a desired outcome in a given situation (Bandura 1989). As a key factor of human agency, SE regulates human functioning via cognitive, motivational, emotional, and selection processes (Bandura 1997, 2000), and generally affects the degree of effort that an individual will expend as well as how long that person will persist in the face of adversity (Bandura 1977). It is well recognized that self-efficacy is positively related to well-being (e.g., Dark-Freudeman and West 2016; Magklara and Morrison 2016; Roos et al. 2013; Shakespeare-Finch et al. 2015; van Seggelen-Damen and van Dam 2016).

1.1 Is Self-Efficacy a General or Specific Construct?

Traditionally, SE has been viewed as “a situation-specific competence belief” (Scherbaum et al. 2006, p. 1047). To capture the specificity of SE and its dependence on a particular situation or context, SE must be tailored to the specific domain of functioning of interest (i.e., specific SE) (Bandura 2006). SE has also been measured as a more general and relatively stable sense of personal competence that applies to, or generalizes across, a variety of different situations (i.e., general SE) (Judge et al. 1998). Sherer et al. (1982) suggest that an individual should have a general set of expectations that is carried to new situations and influences an individual’s mastery expectations in new situations.

Of great debate is whether general SE even makes sense. Washington and Moxley (2013) suggest general and specific SE are both viable; they argue that general SE can help inspire and catalyze change, while specific SE domains plays a role later as individuals move through change processes. Others argue that, whereas general SE might explain a broad range of behaviors and coping mechanisms, particularly in new situations, the majority of applications of SE need to be conceptualized in a situation or task specific manner (Bandura 2006; Eller et al. 2016; Feldman and Kubota 2015; Luszczynska et al. 2005; Sherer et al. 1982; van der Bijl and Shortridge-Baggett 2001). At present, both general and specific conceptions and measures of SE exist.

Assessing SE both generally and in relation to specific areas can provide different information and potentially serve different purposes. It is worth noting that Bandura (1997) claimed general SE measures have little to no relation to specific SE. Bonetti et al. (2001) reported a correlation between a general SE measure and a health-specific SE measure of .38 for a student sample and correlations ranging from .16 to .30 in patient samples. General SE is not intended to be a replacement for specific SE. As Chen et al. (2001) note, general SE “is a supplement that is predicted to be useful when the performance under scrutiny is generalized” (p. 65) whereas specific SE “is a proximal state that positively relates to individuals’ decisions to engage and persist in task-related behavior” (p. 67). Certainly evidence seems to suggest that general SE predicts general performance best whereas specific SE predicts domain specific performance best (see Chen et al. 2001).

1.2 Self-Efficacy in Studies of Homelessness

The focus of much research in the area of homelessness is about health outcomes and the availability and accessibility of health and housing resources (Benston 2015; Brown et al. 2015; O’Brien et al. 2015; Rowe et al. 2016; Su et al. 2015). What is relatively absent from this research is the role that positive psychology and SE, in particular, may play in homeless individuals’ abilities to utilize resources, stay motivated, and persevere in spite of what are often daily challenges. Despite the recent assertion that SE is an important construct in understanding, explaining, and changing health behaviour (de Vries 2016), SE remains an understudied personal variable in research with homeless samples.

There is some research that has examined SE with individuals who are homeless. Homelessness, and even self-categorization as homeless, is linked to lower well-being, feeling powerless, having low SE, and feeling a lack of control over one’s life (Johnstone et al. 2015; Sadow and Hopkins 1993; Walter et al. 2015). Parker et al. (2016) found that the more central or important being homeless was to one’s identity, the lower one’s SE whereas homeless identity salience had no relationship to one’s SE. Guarnaccia and Henderson (1993) found SE was unrelated to age and length of time spent homeless; they used Sherer et al.’s (1982) Self-Efficacy Scale and questioned whether this measure was appropriate and “specific to the ecological realities and circumstances of homelessness” (p. 337). As noted by Washington and Moxley (2013), the “sheer stress of daily living, exposure to degraded environments and lack of opportunity for self-improvement can diminish optimism and hope, as well as reduce motivation for self-help and care” (p. 45). SE can operate as a buffer for daily stress (Schönfeld et al. 2016) and depressive symptoms (Maciejewski et al. 2000).

In research with adolescents, Maccio and Schuler (2012) found that SE scores, measured using a version of Pearlin and Schooler’s (1978) Mastery Scale, were lower in a sample of homeless and runaway youth than in a general adolescent group and were significantly related to self-esteem, but not to alcohol or drug use. In a study by Broadhead-Fearn and White (2006), SE, as measured by three items specifically developed for the study, was the strongest predictor of intentions and the only significant predictor of rule-following behaviour in youth at a homeless youth shelter.

With adults, Epel et al. (1999) used measures of housing- and job-specific SE developed specifically for the study to determine whether these variables affected coping strategies related to obtaining housing and employment at a family shelter. Those with high SE searched more for housing and employment and they stayed at the shelter for a shorter period. Adults with low SE were more likely to request an extension of their stay at the shelter. Epel et al. (1999) described SE as an essential factor related to one’s ability to escape homelessness, manage health related behaviours and reintegrate into society. Also, Seilheimer and Doyal (1996) found that SE was a key predictor of housing satisfaction among individuals with mental health issues in different housing situations.

Generally, the role that SE plays in the use of health and health-related supports, as well as housing transitions and housing quality, among individuals who are homeless or vulnerably housed is poorly understood and may be negatively affected by the lack of consistency in SE measures used (i.e., most measures used were developed specifically for a given study and, therefore, are not widely available) and the fact that many SE measures have not been evaluated psychometrically, particularly with homeless samples.

1.3 Gathering and Evaluating Reliability and Validity Evidence

In order to better evaluate the role of SE with individuals who are homeless or vulnerably housed, it is important to identify relevant and appropriate SE measures that produce reliable scores and permit valid inferences from those scores for this target population. If evidence obtained using a measure in a particular context with a particular sample suggests a poor degree of reliability or validity, then researchers, service providers, and policy makers may be erroneous in their thinking about the role of SE. Specifically, they may misidentify the construct that is of most interest or effectiveness in their work, make erroneous conclusions about relationships among variables, and make poorly informed decisions about the impact of interventions or policy. A lack, or relative lack, of reliability or validity evidence for different SE measures used with the homeless or vulnerably housed individuals means that researchers, service providers, and policy makers are working in the dark and have no idea whether they are measuring the intended construct or making informed decisions based on the intended construct. In short, it is important to (a) identify some relevant and appropriate general and specific SE measures that have the potential to be used across different studies of homeless or vulnerably housed adults, and (b) begin gathering evidence related to the reliability of scores and validity of inferences made from these SE measures.

Three scales measuring SE that may be relevant to individuals who are homeless or vulnerably housed include the Generalized Self-Efficacy Scale (GSES) (Schwarzer and Jerusalem 1995), Perceived Health Competence Scale (PHCS) (Smith et al. 1995), and Housing Self-Efficacy Scale (HSES) (Hubley 2012). These represent one general SE measure and two specific SE measures – one related to health-specific SE and one related to housing-specific SE. These three variables (general SE, health-specific SE, and housing-specific SE) were selected because they are expected to be of greatest interest to researchers and service providers given the focus of homelessness research on health outcomes as well as the availability and accessibility of health and housing resources (Benston 2015; Brown et al. 2015; O’Brien et al. 2015; Rowe et al. 2016; Su et al. 2015).

The Standards for Educational and Psychological Testing1 (AERA, APA, & NCME 1999, 2014) identify five sources of validity evidence that might be used in evaluating the proposed inferences from test scores obtained in particular contexts with different samples: (1) test content using content experts, (2) internal structure using factor analyses, (3) relations to other variables (which includes convergent/discriminant evidence, test-criterion relationships), (4) response processes, and (5) consequences of testing. As noted in the Standards (p. 14), not all evidence is applicable with all constructs or with all uses of a measure. One source of evidence that is foundational is internal structure using factor analysis as this not only determines if the dimensionality of the measure is in line with what is expected theoretically or empirically but also informs how a measure should be scored to appropriately capture the variability in the participants’ responses to scale items (Flora and Flake 2017; Hubley and Zumbo 2013; Slocum-Gori and Zumbo 2011; Slocum-Gori et al. 2009; Zumbo 2007, 2009). Next, it is appropriate to provide a reliability estimate for any scores as reliability impacts the magnitude of effects in subsequent statistical analyses (e.g., correlation – including validity coefficients, analyses of variance, regression) (Shear and Zumbo 2013; Zimmerman and Zumbo 2015).

The GSES is expected to be unidimensional based on past research (Scholz et al. 2002; Schwarzer and Born 1997), including a cross-national study comparing the factor structure in 28 countries (Scholz et al. 2002). Internal consistency reliability estimates in other samples have ranged from .75 to .91 (Scholz et al. 2002). The PHCS is also expected to be unidimensional based on a study that reported a one-factor structure with patients from a private practice (Dempster and Donnelly 2008). While Bonetti et al. (2001) reported PHCS items loaded on one factor for students, they found the items loaded on two factors (reflecting positively and negatively worded items) for patients. Dempster and Donnelly (2008) reported an internal consistency reliability estimate of .91 for the PHCS. There is surprisingly little psychometric evidence reported for English language versions of the GSES and PHCS. The HSES is newly developed and no internal structure analyses have been conducted with this measure. Thus, at present, there appears to be no published research addressing the factor structure of the GSES, PHCS, or HSES, or the reliability of scores obtained from these measures, with homeless or vulnerably housed adults.

1.4 Study Purpose

The purposes of this study were: (a) to examine whether the item response data from a group of homeless or vulnerably housed adults for each of the GSES, PHCS and HSES fit a strictly unidimensional factor model, (b) if strict unidimensionality for a scale was not supported, to examine whether the item response data from each scale fits essential unidimensionality or some multidimensional structure, (c) based on these results, indicate how to best score each SE measure and then obtain internal consistency reliability estimates for the scores, (d) report the inter-correlations among the three SE measures, (e) report descriptive performance (mean, SD, skewness, kurtosis) for each measure, and (f) examine the role of demographic variables with each of the three SE measures.

2 Methods

2.1 Participants

Data from 323 individuals were taken from the Vancouver cohort of the Canadian Institutes for Health Research funded Health and Housing in Transition longitudinal study (Hwang et al. 2011) in 2012 (year 4), when data for all three SE measures were collected. Only participants with complete data were included given the small number of missing data (n = 8; 2.4%). Participants were identified at baseline as (a) homeless if they did not have their own place and were currently living in a public place, shelter, vehicle, abandoned building, or someone else’s place without paying rent (n = 159; 49.2%), or (b) vulnerably housed if they reported living in their own apartment, room, or place and were paying rent and also had been homeless in the past 12 months or had two or more moves in the past 12 months (n = 164; 50.8%).2 Homeless adults were recruited from both shelters and meal programs whereas vulnerably housed participants were recruited from single room occupancy hotels, meal programs, community health care centers, and drop-in centers. The sampling methods used for recruitment are described in detail elsewhere (Hwang et al. 2011).

The sample consisted of 203 men (62.8%), 114 women (35.3%), and 6 individuals who identified as transgendered (1.9%). At the time of data collection for this study (year 4 - when the SE data were collected), participants ranged in age from 22 to 75 years (M = 45.0, SD = 10.04). Participants identified as: White (n = 185; 57.3%), First Nations/Aboriginal (n = 85; 26.3%), of mixed ethnicity (n = 24; 7.4%), or Black/African Canadian (n = 10; 3.1%). In terms of education, 147 (45.5%) had some high school, 78 (24.1%) had completed high school, and 94 (29.1%) had some post-secondary education or higher. At this point in the study, 268 (83.0%) of the participants were vulnerably housed, 41 (12.7%) were homeless, and 12 (3.7%) were in institutions. In the previous 12 months, 135 (41.8%) of the sample had worked at a paid job and 58 (18%) had been incarcerated.

2.2 Procedures

Data were collected using structured in-person interviews that lasted between 60 to 90 minutes (Hwang et al. 2011). While a variety of variables, including, for example, demographic characteristics, quality of life, housing quality, and physical and mental health status, were collected, it is primarily data from the three SE measures that are of interest in this particular study. Study participants provided written informed consent and were compensated $20 CAD. The study was approved by the institutional research ethics board of the researchers.

2.3 Measures

The following three SE scales measuring general, health-specific, and housing-specific SE, respectively, were used.

General Self Efficacy Scale (GSES)

The GSES (Schwarzer and Jerusalem 1995) is a 10-item measure that assesses a general sense of perceived self-efficacy or optimistic self-belief in one’s ability to respond to novel or difficult situations. Participants were asked to respond using a 4-point response scale, ranging from 1 (not true at all) to 4 (completely true). A total score is obtained by summing the responses to all 10 items and ranges from 10 to 40; higher scores reflect greater general SE.

Perceived Health Competence Scale (PHCS)

The PHCS (Smith et al. 1995) is an 8-item measure that examines the degree to which an individual feels able to manage health-related behaviors and outcomes across a range of situations. Participants were asked to rate their level of agreement with each item using a 5-point scale, which ranged from 1 (strongly disagree) to 5 (strongly agree). A total score is computed by summing the responses to all 8 items and ranges from 8 to 40; higher scores reflect greater health-specific SE.

Housing Self-Efficacy Scale (HSES)

The HSES (Hubley 2012) is a 7-item measure that examines whether an individual believes he/she can do what is needed to improve or manage his/her housing situation. The HSES was developed specifically for use in the Health and Housing in Transition study (Hwang et al. 2011). Participants responded to items using a 5-point scale, ranging from 1 (strongly disagree) to 5 (strongly agree). A total score is computed by summing the responses to all 7 items and ranges from 7 to 35; higher scores reflect greater housing-specific SE.

2.4 Statistical Analyses

Internal Structure

Strict unidimensionality was examined by conducting separate one-factor confirmatory factor analyses for the three SE measures using the package ‘lavaan’ (Rosseel 2012) from the freely available software R. Given the number of response options for the GSES (4-point), PHCS (5-point), and HSES (5-point), the item response data from all three measures were treated as ordinal, and polychoric correlation matrices were used (Finney and DiStefano 2013; Kaplan 2009; Zumbo et al. 2002). Robust diagonally weighted least squares estimation was used, as recommended with ordinal data (Bandalos 2014; Finney and DiStefano 2013). Hu and Bentler’s (1999) combination rule for two fit indices, Comparative Fit Index ≥ .95 and Root Mean Square Error of Approximation < .08, were used. We also took into account the Tucker-Lewis Index (> .95) and the chi-square test (χ2; looking for a statistically non-significant finding). Importantly, we also examined the residual polychoric correlation matrices to evaluate model fit. Values within ±0.10 are considered adequate in residual correlation matrices (McDonald 1985; Zumbo and Taylor 1993) but values above this criterion indicate a lack of strict unidimensionality.

If, based on the confirmatory factor analyses, a strictly unidimensional model was rejected, then the data were evaluated for essential unidimensionality using exploratory factor analysis using minres estimation and oblimin rotation. Essential unidimensionality was identified by examining the ratio of first to second eigenvalues; if the ratio is greater than 4.0, there is evidence for essential unidimensionality (Slocum-Gori and Zumbo 2011). If evidence does not support essential unidimensionality, then multidimensionality was considered and parallel analysis was used to determine the number of factors, using the package ‘psych’ from R (Revelle 2017). Factor loadings ≥ .40 were set as the criterion for an item to load on a given factor.

Scoring of the GSES, PHCS, and HSES

Scoring of the three SE measures was based on the internal structure results for each measure. It is critical that internal structure inform scoring, although this is a step that researchers often ignore. Strict unidimensionality from confirmatory factor analyses indicates all items load on only one latent variable, and it provides a clear indication that a total score can be used (Slocum-Gori et al. 2009). Essential unidimensionality indicates there is one dominant factor and other minor (but not additional) factor(s) as determined from an exploratory factor analysis, and is also sufficient to allow the use of a total score (Slocum-Gori et al. 2009; Slocum-Gori and Zumbo 2011). The presence of a multidimensional internal structure would support the use of multiple subscale scores rather than a total score.

Score Reliability Estimation

Given that the item responses were treated as ordinal, score reliability was examined using ordinal alpha (Gadermann et al. 2012; Zumbo et al. 2007) using the package ‘psych’ from R (Revelle 2017). Ordinal alpha >.80 was considered satisfactory.

Scale Inter-Correlations and Relationships with Demographic Variables

The relationships among the scores on the GSES, PHCS, and HSES as well as between age and scores on each of the three SE measures were examined using Pearson correlations. Gender and education differences on each of the three SE measures were examined using separate 2 (male, female) × 3 (educational level) between-groups analyses of variance. To control for experiment-wise error, the Type I error rate for each analysis of variance was adjusted using a Bonferroni correction (i.e., α = .017 or .05/3) as per Huberty and Morris (1989).

3 Results

3.1 Strict Unidimensionality of the Self-Efficacy Scales

The polychoric correlations ranged from .33 to .73 for the GSES, from .25 to .71 for the PHCS, and from .28 to .63 for the HSES, supporting their appropriateness for factor analysis. One factor confirmatory factor analysis models using robust diagonally weighted least squares estimation utilizing a polychoric matrix did not show acceptable model fit for strict unidimensionality for the GSES, PHCS or HSES (see Table 1). Comparative Fit Index and Tucker-Lewis Index values were satisfactory for the GSES and HSES, but not the PHCS. Chi-square values were all statistically significant and Root Mean Square Error of Approximation values and their 90% confidence intervals, were well over the criterion cut-off for all scales.
Table 1

Assessing strict unidimensionality of the three self-efficacy scales

 

χ2 Test and goodness of fit indices

χ2 (df)

Comparative fit index (CFI)

Tucker-Lewis index (TLI)

Root mean square error of approximation (RMSEA)

RMSEA 90% confidence interval

General Self Efficacy Scale

181.48** (35)

.969

.960

.114

.098 – .131

Perceived Health Competence Scale

236.06** (20)

.916

.883

.183

.163 – .204

Housing Self-Efficacy Scale

81.56** (14)

.971

.956

.122

.097 – .149

** p < .001

We also examined closely the residual polychoric correlation matrices. This examination showed 1 problematic residual correlation (i.e., greater than ±.10) for the GSES (out of 55; 1.8%), 5 problematic residual correlations for the PHCS (out of 36; 13.9%), and 1 problematic residual correlation for the HSES (out of 28; 3.6%) (see Table 2).
Table 2

Residual polychoric correlation matrices for the three self-efficacy scales

General Self Efficacy Scale

 Item

1

2

3

4

5

6

7

8

9

10

 1

.00

         

 2

.10

.00

        

 3

.10

.08

.00

       

 4

−.01

.00

.02

.00

      

 5

−.03

−.03

−.01

.06

.00

     

 6

.02

−.04

−.00

.02

.04

.00

    

 7

−.09

−.10

−.08

−.02

.00

−.05

.00

   

 8

−.07

−.02

−.07

−.04

−.08

−.03

.09

.00

  

 9

−.06

−.02

−.13

−.07

−.09

−.01

.05

.08

.00

 

 10

−.05

−.05

.01

−.04

.02

−.05

.04

.00

.06

.00

Perceived Health Competence Scale

 Item

1

2

3

4

5

6

7

8

  

 1

.00

         

 2

−.10

.00

        

 3

−.07

.11

.00

       

 4

−.04

−.17

−.09

.00

      

 5

.08

−.14

−.10

.13

.00

     

 6

−.01

.01

.01

−.06

−.05

.00

    

 7

−.01

.07

.01

−.07

−.10

.07

.00

   

 8

.14

−.09

−.02

.05

.10

−.10

−.07

.00

  

Housing Self-Efficacy Scale

 Item

1

2

3

4

5

6

7

   

 1

.00

         

 2

−.01

.00

        

 3

−.02

−.02

.00

       

 4

.11

−.01

−.03

.00

      

 5

−.03

.02

.02

−.06

.00

     

 6

−.01

.05

−.05

.06

−.10

.00

    

 7

−.10

−.00

.09

−.10

.04

.01

.00

   

Residual values >|.10| are bolded

3.2 Essential Unidimensionality of the Self-Efficacy Scales

The ratio of first to second eigenvalues of the polychoric correlation matrix obtained using exploratory factor analysis showed values well above the criterion (> 4.0) for all scales: GSES = 14.7, PHCS = 7.4, HSES = 13.3; this strongly supports the essential unidimensionality of each SE scale. For each SE measure, all items loaded > .40 for a one-factor solution (see Table 3).
Table 3

Single factor loadings for the three self-efficacy scales

Item

General Self Efficacy Scale (10 items)

Item

Perceived Health Competence scale (8 items)

Item

Housing Self-Efficacy Scale (7 items)

1

.768

1

.482

1

.714

2

.612

2

.724

2

.799

3

.777

3

.724

3

.777

4

.822

4

.635

4

.660

5

.804

5

.802

5

.764

6

.863

6

.674

6

.499

7

.706

7

.876

7

.690

8

.807

8

.610

  

9

.796

    

10

.765

    

3.3 Scoring and Reliability Estimates

Given that essential unidimensionality was supported for each of the GSES, PHCS, and HSES, item responses can be summed to create a total score for each measure. Ordinal alpha coefficients were .93 for the GSES, .87 for the PHCS, and .87 for the HSES. An examination of ordinal alpha-if-item-deleted values showed that the reliability estimates would not improve if any of the items were deleted from each of the measures.

3.4 Performance on the Three Self-Efficacy Measures and Scale Inter-Correlations

Performance on the GSES, PHCS, and HSES is presented in Table 4. The GSES scores showed low moderate correlations with both PHCS scores (r = .36, p < .01) and HSES scores (r = .39, p < .01). Scores on the two specific SE scales, PHCS and HSES, showed a moderate, but higher, correlation with each other, at r = .45 (p < .01).
Table 4

Performance on the three self-efficacy measures

 

Theoretical range

Obtained range

Mean

Standard deviation

Skewness

Kurtosis

General Self Efficacy Scale

10 – 40

12 – 40

29.59

6.53

−0.32

−0.46

Perceived Health Competence Scale

8 – 40

13 – 40

27.46

5.50

−0.34

−0.60

Housing Self-Efficacy Scale

7 – 35

10 – 35

23.98

5.30

−0.34

−0.68

3.5 Relationships between the Three Self-Efficacy Measures and Demographic Variables

No statistically significant relationship was found between age and either GSES (r = −.07, n.s.) or PHCS (r = −.02, n.s.) but there was a small significant relationship between age and HSES (r = −.14, p < .05).

Gender by educational level differences for each of the three SE measures were examined by conducting 2 × 3 between-groups analyses of variance. The results of the omnibus F-tests are presented in Table 5. There were no significant main effects for gender on any of the SE measures. There were no statistically significant main effects for educational level on the specific SE measures (PHCS and HSES), but there was a small statistically significant main effect for educational level on the GSES. Tukey’s Honestly Significant Difference post-hoc tests indicated that individuals with only some high school (n = 144, M = 28.25, SD = 6.56) scored significantly lower in general self-efficacy than those who completed high school (n = 77, M = 30.88, SD = 6.01) or had some post-secondary schooling or higher (n = 92, M = 30.61, SD = 6.49). There were no significant gender by educational level interactions on any of the SE measures.
Table 5

Gender x education analysis of variance results for each of the three self-efficacy measures

Self-efficacy measure

Omnibus F-test results

Effect size p2)

General Self Efficacy Scale

 Main effect of gender

F(1, 307) = 0.81, n.s.

.003

 Main effect of education

F (2, 307) = 5.17**

.033

 Gender x education interaction

F(2, 307) = 0.35, n.s.

.002

Perceived Health Competence Scale

 Main effect of gender

F(1, 307) = 0.74, n.s.

.002

 Main effect of education

F(2, 307) = 1.36, n.s.

.009

 Gender x education interaction

F(2, 307) = 0.53, n.s.

.003

Housing Self-Efficacy Scale

 Main effect of gender

F(1, 307) = 0.09, n.s.

.000

 Main effect of education

F(2, 307) = 0.49, n.s.

.003

 Gender x education interaction

F(2, 307) = 0.05, n.s.

.000

** p < .01. ηp2 = partial eta squared (effect size). Significant results are bolded (α was set to .017)

4 Discussion

Limited research has been conducted that examines the role of, or makes use of information about, general or specific SE in the lives of individuals who are homeless or vulnerably housed. SE is a belief that one can engage in a course of action to bring about a desired outcome. It can provide information about one’s perceived ability to accomplish difficult tasks by serving as a mediating variable between challenging situations and desired outcomes (Washington and Moxley 2003). For individuals who are homeless or vulnerably housed, SE may (a) serve an important role in understanding whether individuals will access and use resources (e.g., to obtain more stable housing, improve their health status), (b) help in attempts to build on individuals’ resilience and either maintain or strengthen one’s SE, and (c) suggest how service providers can enhance the environment that such individuals must navigate in order to achieve their goals. In good part, research is limited because there are few measures that have either been used in common across studies or have had reliability of test scores or validity of inferences examined. While general SE is broad in scope, specific SE focuses on particular domains of one’s life (e.g., health, housing). We selected three different measures of SE (one general SE and two specific SE – one for health and one for housing) that have the potential to provide different types of information about SE of particular relevance to research and service provision for individuals who are homeless or vulnerably housed. This study fills a large measurement gap in that there is no information about the internal structure of these measures with homeless or vulnerably housed adults and limited psychometric evidence with other samples.

The primary purpose of the present study was to examine internal structure validity evidence, scoring, and reliability estimates of scores for the GSES, PHCS, and HSES in a sample of individuals who are homeless or vulnerably housed. Internal structure validation studies are often not conducted because they require relatively large sample sizes (typically N > 200) but evidence is needed about the internal structure of a measure before one can determine how to appropriately score the measure, estimate the reliability of any scores, or examine other validity evidence. We also examined the inter-correlations among the scores on these three SE measures, described performance on each of the measures, and reported correlations with age and any differences based on gender or educational level.

In this study, we found that none of the SE measures met the criteria for strict unidimensionality, but all three measures clearly exceeded the criterion for essential unidimensionality. It is generally theoretically expected that SE measures are unidimensional (Dempster and Donnelly 2008; Scholz et al. 2002; Schwarzer et al. 1999); the results of this study are consistent with this expectation and provide evidence based on the internal structure source for validity (AERA et al. 1999, 2014). In addition, this finding supports the use of total scores to represent the responses of an adult sample of individuals who were homeless or vulnerably housed to items on each of the GSES, PHCS, and HSES. Inter-correlations among the three measures ranged from .36 to .45, suggesting relatively little shared variance. While these inter-correlations are somewhat consistent with theory (Bandura 1997) and previous empirical work, they are a little higher than reported correlations of .16 to .38 between general SE and health-specific SE measures for different samples (e.g., Bonetti et al. 2001). Ordinal alpha coefficients were .93 for GSES scores, .87 for PHCS scores, and .87 for HSES scores. These reliability estimates all clearly exceeded the .80 criterion, ranging from very good to excellent. They are also consistent with previous research reporting reliability estimates obtained with a variety of samples, ranging mostly from .87 to .95 for the GSES (Nilsson et al. 2015; Peter et al. 2014; Ponizovsky et al. 2011; Scholz et al. 2002; Schwarzer et al. 1999) and from .82 to .91 for the PHCS (Dempster and Donnelly 2008; Keefer et al. 2014; Padden et al. 2013; Smith et al. 1995).

We also examined the relationship of age, gender, and education to GSES, PHCS, and HSES scores. In the present study, relationships between age and both GSES and PHCS were near-zero whereas the correlation between age and HSES was significant but small (r = −.14). Guarnaccia and Henderson (1993) also reported a very low, nonsignificant correlation between age and a measure of general self-efficacy. Scholz et al. (2002) suggested that a lack of age effects is consistent with SE theory. There were no significant gender by education interactions or main effects of gender on any of the SE measures, but there was a small statistically significant main effect for education only on general SE, with individuals with lower educational level showing significantly lower SE than those with higher education levels. Epel et al. (1999) also reported nonsignificant relationships between educational level and domain-specific SE in a sample of homeless men and women. Given the limited literature on the relationship of demographic variables to SE, it is difficult to assess the significance of these findings.

Obtained and possible score ranges for each of the three SE measures were very close; moreover, the low skewness and kurtosis values suggest fairly normal distributions. The mean GSES score in the sample was 29.6 (SD = 6.5), which is similar to the mean scores found with other samples, including university students and young adults (M = 30.5–31.5; Luszczynska et al. 2005; Schwarzer et al. 1999; Wu 2009), individuals with spinal cord injury (M = 31.6; Peter et al. 2014), patients with Parkinson’s disease (M = 28.4; Nilsson et al. 2015), and adults (M = 28.6–29.5; Ponizovsky et al. 2011; Wu 2009), but higher than that found with patients with adjustment disorder (M = 21.8; Ponizovsky et al. 2011). The mean PHCS score in our study was 27.5 (SD = 5.5), which is lower than that found with different patient samples (M = 28.0–29.9; Dempster and Donnelly 2008; Keefer et al. 2014; Smith et al. 1995) and non-patient groups (M = 30.1–32.4; Padden et al. 2013; Smith et al. 1995). The mean HSES score in the present study was 24.0; however, as this is a new measure, there is no prior research to which we can compare our findings. Nonetheless, one must be cautious in making comparisons to other samples without evidence of measurement invariance (MI). MI suggests that associations between item scores and latent variables do not vary based on group membership. Thus, a measure can be deemed to be measuring the same attribute across groups only once MI has been shown. There are different levels of MI, with configural invariance needed to assume the factor patterns are the same across subgroups, metric (weak) invariance needed to examine structural relationships or correlations using a measure across groups, and scalar (strong) invariance needed to examine mean differences across groups (Horn and McArdle 1992; Steenkamp and Baumgartner 1998). Notably, without evidence of scalar invariance, any differences found between groups cannot be interpreted unambiguously to mean the groups differ on the same construct.

4.1 Study Limitations and Future Research

A limitation of the present study is that we were unable to examine MI across specific groups within our own sample. For example, the sample sizes for men and women were not large enough to examine measurement invariance; thus it remains unclear whether one can assume that men and women think about general, health-specific, and housing-specific SE in the same way and whether one can meaningfully make comparisons of correlational evidence or mean differences between men and women using GSES, PHCS, and HSES scores. It would be beneficial to examine this in future research.

Understandably, there may be interest in comparing homeless and vulnerably housed subgroups using each of the SE measures. Roughly equal numbers of homeless and vulnerably housed individuals were recruited in the larger study at baseline. Data on the three SE measures, however, were not collected until year 4. Over the course of the longitudinal study, it was discovered that many participants cycled between homeless and vulnerably housed states over the course of a year or across years (e.g., given fluctuations in mental health, substance use, employment) and thus these groups likely reflect states of being rather than distinct populations as had been envisioned initially. In fact, at the time of the year 4 interview, 83.5% of the sample was vulnerably housed and only 12.8% was homeless (even though decreasing homelessness was not a component or purpose of the larger longitudinal study). As a result, in the present study, we determined that, even if the sample sizes had been large enough, examining MI between homeless and vulnerably housed groups on the three SE measures would not have been a meaningful exercise. Still, there may be other ways to meaningfully categorize participants (e.g., based on length of homelessness or stability of homeless/housed state) in future research and such studies may benefit from examining MI in SE among those groups if they wish to treat the groups separately or compare them. In addition, researchers wanting to compare SE in a homeless or vulnerably housed group to a control group (e.g., individuals who are stably housed) should also examine MI on SE measures between these groups.

With respect to the measurement and use of SE, it is interesting that there has been little to no research on the relationship between psychological well-being and SE in homeless or vulnerably housed adults. The very components of psychological well-being (i.e., whether individuals’ lives have purpose, whether they are realizing their potential, if they feel in charge of their own lives, and what is the quality of their relationships with others (Ryff and Keyes 1995) and the roles that general and specific SE might play would seem of particular interest with this population. It is hoped that the availability of both general and specific SE scales that have some preliminary evidence to support their use with homeless and vulnerably housed individuals will lead to more studies, with greater measurement rigour, that examine the role that SE plays in the daily lives of individuals who are homeless or vulnerably housed as well as in their use of health and housing supports. Notably, specific SE measures are likely to be more highly predictive when used to predict domain or situation specific outcomes. For example, health-specific SE is likely to be more highly predictive when the outcomes are health-related whereas housing-specific SE is likely to be more highly predictive when the outcomes are housing-related. It is expected that general SE, which would generally reflect one’s past experiences, would play a stronger role in novel situations in which the individual has little information or specific experience (Sherer et al. 1982). While somewhat promising, this type of research is still in its infancy in the field of homelessness research and is limited by a lack of measures that have been properly examined with this target population.

A second limitation of the present study is that we were unable to examine other sources of validity evidence (e.g., test content validity, relations with other variables) as this was not a goal of the larger study. As noted by Downing (2003), “validity is never assumed and is an ongoing process of hypothesis generation, data collection and testing, critical evaluation and logical inference” (p. 831). Thus, future research should address other sources of validity evidence (AERA et al. 2014) in evaluating the GSES, PHCS, and HSES when used with individuals who are homeless or vulnerably housed. In particular, future research should address: (a) test content to ensure that the intended construct of SE is being measured in line with theory, (b) response processes to ensure the target population is responding to the items in ways that are expected if the construct measured is SE or if there is evidence of additional or alternative constructs being measured (see, for example, Zumbo and Hubley 2017), and (c) relations with other variables to determine if the pattern of correlations found with convergent and discriminant measures is consistent with that expected in the nomological network associated with SE (as opposed to alternative or competing inferences, such as motivation or self-esteem) (see, for example, Bandura 1986; Schwarzer and McAuley 2016; Sheer 2014; Williams and Rhodes 2016; Zimmerman 2000). Eventually, given the strong policy and intervention implications of SE findings for the homeless and vulnerably housed population, evidence related to test consequences should be examined to determine the intended consequences and unintended side effects (Hubley and Zumbo 2011; Zumbo and Hubley 2016) of legitimate SE test score interpretation and use.

Footnotes

  1. 1.

    Henceforth referred to as the Standards.

  2. 2.

    Over the course of the longitudinal study, many participants cycled between homeless and vulnerably housed states, thus this status is more relevant to the recruitment process.

Notes

Acknowledgements

This work was supported by the Canadian Institute for Health Research (CIHR) under operating grant MOP-86765 awarded to Drs. Stephen Hwang, Tim Aubry, Susan Farrell, Anita Palepu, James Dunn, Anita Hubley, Jeffrey Hoch, J. David Hulchanski, Fran Klodawsky, and Rosane Nisenbaum and under an Interdisciplinary Capacity Enhancement Grant on Homelessness, Housing and Health (HOA-80066) awarded to Drs. Stephen Hwang, Tim Aubry, Anita Palepu, Anita Hubley, James Dunn, Jeffrey Hoch, J. David Hulchanski, Bruce MacLaurin, Elise Roy, Jeffrey Turnbull, and Catherine Worthington.

Compliance with Ethical Standards

Conflict of Interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

References

  1. American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. [AERA, APA, & NCME]. (1999). Standards for educational and psychological testing. Washington, DC: American Educational Research Association.Google Scholar
  2. American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. [AERA, APA, & NCME]. (2014). Standards for educational and psychological testing. Washington, DC: American Educational Research Association.Google Scholar
  3. Bandalos, D. L. (2014). Relative performance of categorical diagonally weighted least squares and robust maximum likelihood estimation. Structural Equation Modeling, 21, 102–116.  https://doi.org/10.1080/10705511.2014.859510.CrossRefGoogle Scholar
  4. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191–215.CrossRefGoogle Scholar
  5. Bandura, A. (1986). The explanatory and predictive scope of self-efficacy theory. Journal of Social and Clinical Psychology, 4, 359–373.  https://doi.org/10.1521/jscp.1986.4.3.359.CrossRefGoogle Scholar
  6. Bandura, A. (1989). Human agency in social cognitive theory. The American Psychologist, 44, 1175–1184.CrossRefGoogle Scholar
  7. Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman and Company.Google Scholar
  8. Bandura, A. (2000). Self-efficacy: The foundation of agency. In W. J. Perrig, A. Grob, W. J. Perrig, & A. Grob (Eds.), Control of human behavior, mental processes, and consciousness: Essays in honor of the 60th birthday of august Flammer (pp. 17–33). Mahwah: Lawrence Erlbaum Associates Publishers.Google Scholar
  9. Bandura, A. (2006). Guide for constructing self-efficacy scales. In F. Pajares & T. C. Urdan (Eds.), Self-efficacy beliefs of adolescents (pp. 307–337). USA: Information Age Publishing.  https://doi.org/10.1017/CBO9781107415324.004.Google Scholar
  10. Benston, E. A. (2015). Housing programs for homeless individuals with mental illness: Effects on housing and mental health outcomes. Psychiatric Services, 66, 806–816.  https://doi.org/10.1176/appi.ps.201400294.CrossRefGoogle Scholar
  11. van der Bijl, J. J., & Shortridge-Baggett, L. M. (2001). The theory and measurement of the self-efficacy construct. Scholarly Inquiry for Nursing Practice, 15, 189–207.Google Scholar
  12. Bonetti, D., Johnston, M., Rodriguez-marin, J., Pastor, M., Martin-aragon, M., Doherty, E., & Sheehan, K. (2001). Dimensions of perceived control: A factor analysis of three measures and an examination of their relation to activity level and mood in a student and cross-cultural patient sample. Psychology & Health, 16, 655-674,  https://doi.org/10.1080/08870440108405865.
  13. Bornstein, M. H., Davidson, L., Keyes, C. L. M., & Moore, K. A. (2003). Well-being: Positive development across the life course. Mahwah: Lawrence Erlbaum Associates.Google Scholar
  14. Broadhead-Fearn, D., & White, K. M. (2006). The role of self-efficacy in predicting rule-following behaviors in homeless youth: A test of the theory of planned behavior. Journal of Social Psychology, 146, 307–325.CrossRefGoogle Scholar
  15. Brown, R. T., Miao, Y., Mitchell, S. L., Bharel, M., Patel, M., Ard, K. L., et al. (2015). Health outcomes of obtaining housing among older homeless adults. American Journal of Public Health, 105, 1482–1488.  https://doi.org/10.2105/AJPH.2014.302539.CrossRefGoogle Scholar
  16. Chen, G., Gully, S. M., & Eden, D. (2001). Validation of a new general self-efficacy scale. Organizational Research Methods, 4, 62–83.CrossRefGoogle Scholar
  17. Dark-Freudeman, A., & West, R. L. (2016). Possible selves and self-regulatory beliefs: Exploring the relationship between health selves, health efficacy, and psychological well-being. The International Journal of Aging & Human Development, 82, 139–165.  https://doi.org/10.1177/0091415015627666.CrossRefGoogle Scholar
  18. de Vries, H. (2016). Self-efficacy: Skip the main factor paradigm! A comment on Williams and Rhodes (2016). Health Psychology Review, 10, 140–143.  https://doi.org/10.1080/17437199.2016.1163234.CrossRefGoogle Scholar
  19. Dempster, M., & Donnelly, M. (2008). Validity of the perceived health competence scale in a UK primary care setting. Psychology, Health & Medicine, 13, 123–127.  https://doi.org/10.1080/13548500701351984.CrossRefGoogle Scholar
  20. Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125, 276–302.CrossRefGoogle Scholar
  21. Downing, S. M. (2003). Validity: On the meaningful interpretation of assessment data. Medical Education, 37, 830–837.  https://doi.org/10.1046/j.1365-2923.2003.01594.x.CrossRefGoogle Scholar
  22. Eller, L. S., Lev, E. L., Yuan, C., & Watkins, A. V. (2016). Describing self-care self-efficacy: Definition, measurement, outcomes, and implications. International Journal of Nursing Knowledge.  https://doi.org/10.1111/2047-3095.12143.
  23. Epel, E. S., Bandura, A., & Zimbardo, P. G. (1999). Escaping homelessness: The influences of self-efficacy and time perspective on coping with homelessness. Journal of Applied Social Psychology, 29, 575–596.  https://doi.org/10.1111/j.1559-1816.1999.tb01402.x.CrossRefGoogle Scholar
  24. Feldman, D. B., & Kubota, M. (2015). Hope, self-efficacy, optimism, and academic achievement: Distinguishing constructs and levels of specificity in predicting college grade-point average. Learning and Individual Differences, 37, 210–216.  https://doi.org/10.1016/j.lindif.2014.11.022.CrossRefGoogle Scholar
  25. Finney, S. J., & DiStefano, C. (2013). Nonnormal and categorical data in structural equation modelling. In G. R. Hancock & R. O. Mueller (Eds.), Structural equation modeling: A second course (2nd ed., pp. 439–492). USA: Information Age Publishing.Google Scholar
  26. Flora, D. B., & Flake, J. K. (2017). The purpose and practice of exploratory and confirmatory factor analysis in psychological research: Decisions for scale development and validation. Canadian Journal of Behavioural Science, 49, 78–88.CrossRefGoogle Scholar
  27. Gadermann, A. M., Guhn, M., & Zumbo, B. D. (2012). Estimating ordinal reliability for Likert-type and ordinal item response data: A conceptual, empirical, and practical guide. Practical Assessment, Research & Evaluation, 17, 1–12.Google Scholar
  28. Guarnaccia, V., & Henderson, J. J. (1993). Self-efficacy, interpersonal competence, and social desirability in homeless people. Journal of Community Psychology, 21, 335–338.  https://doi.org/10.1002/1520-6629(199310)21:4<335::AID-JCOP2290210409>3.0.CO;2-O.CrossRefGoogle Scholar
  29. Horn, J. L., & McArdle, J. J. (1992). A practical and theoretical guide to measurement invariance in aging research. Journal of Experimental Aging Research, 18, 117–144.CrossRefGoogle Scholar
  30. Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55.  https://doi.org/10.1080/10705519909540118.CrossRefGoogle Scholar
  31. Huberty, C. J., & Morris, J. D. (1989). Multivariate analysis versus multiple univariate analyses. Psychological Bulletin, 105, 302–308.  https://doi.org/10.1037/0033-2909.105.2.302.CrossRefGoogle Scholar
  32. Hubley, A. M. (2012). Housing self-efficacy scale. Unpublished measure.Google Scholar
  33. Hubley, A. M., & Zumbo, B. D. (2011). Validity and the consequences of test interpretation and use. Social Indicators Research, 103, 219–230.CrossRefGoogle Scholar
  34. Hubley, A. M., & Zumbo, B. D. (2013). Psychometric characteristics of assessment procedures: An overview. In K. F. Geisinger (Ed.), APA handbook of testing and assessment in psychology (Vol. 1, pp. 3–19). Washington, D.C.: American Psychological Association Press.Google Scholar
  35. Hwang, S. W., Aubry, T., Palepu, A., Farrell, S., Nisenbaum, R., Hubley, A. M., et al. (2011). The health and housing in transition study: A longitudinal study of the health of homeless and vulnerably housed adults in three Canadian cities. International Journal of Public Health, 56, 609–623.  https://doi.org/10.1007/s00038-011-0283-3.CrossRefGoogle Scholar
  36. Johnstone, M., Jetten, J., Dingle, G. A., Parsell, C., & Walter, Z. C. (2015). Discrimination and well-being amongst the homeless: The role of multiple group membership. Frontiers in Psychology, 6, article 739.Google Scholar
  37. Judge, T. A., Erez, A., & Bono, J. A. (1998). The power of being positive: The relation between positive self-concept and job performance. Human Performance, 11, 167–187.CrossRefGoogle Scholar
  38. Kaplan, D. (2009). Structural equation modeling: Foundations and extensions (2nd ed.). Thousand Oaks: SAGE Publications.  https://doi.org/10.4135/9781452226576.CrossRefGoogle Scholar
  39. Kashdan, T. B., Biswas-Diener, R., & King, L. (2008). Reconsidering happiness: The costs of distinguishing between hedonics and eudaimonia. The Journal of Positive Psychology, 3, 219–233.CrossRefGoogle Scholar
  40. Keefer, L., Kiebles, J. L., & Taft, T. H. (2014). The role of self-efficacy in inflammatory bowel disease management: Preliminary validation of a disease-specific measure. Inflammatory Bowel Diseases, 17, 614–620.  https://doi.org/10.1002/ibd.21314.CrossRefGoogle Scholar
  41. Luszczynska, A., Gutiérrez-Doña, B., & Schwarzer, R. (2005). General self-efficacy in various domains of human functioning: Evidence from five countries. International Journal of Psychology, 40, 80–89.  https://doi.org/10.1080/00207590444000041.CrossRefGoogle Scholar
  42. Maccio, E. M., & Schuler, J. T. (2012). Substance use, self-esteem, and self-efficacy among homeless and runaway youth in New Orleans. Child and Adolescent Social Work Journal, 29, 123–136.  https://doi.org/10.1007/s10560-011-0249-6.CrossRefGoogle Scholar
  43. Maciejewski, P. K., Prigerson, H. G., & Mazure, C. M. (2000). Self-efficacy as a mediator between stressful life events and depressive symptoms - differences based on history of prior depression. British Journal of Psychiatry, 176, 373–378.  https://doi.org/10.1192/bjp.176.4.373.CrossRefGoogle Scholar
  44. Magklara, E., & Morrison, V. (2016). The associations of illness perceptions and self-efficacy with psychological well-being of patients in preparation for joint replacement surgery. Psychology, Health & Medicine, 21, 735–742.  https://doi.org/10.1080/13548506.2015.1115109.CrossRefGoogle Scholar
  45. McDonald, R. P. (1985). Factor analysis and related methods. Hillsdale: Lawrence Erlbaum Associates Publishers.Google Scholar
  46. Nilsson, M. H., Hagell, P., & Iwarsson, S. (2015). Psychometric properties of the general self-efficacy scale in Parkinson's disease. Acta Neurologica Scandinavica, 132(2), 89–96.  https://doi.org/10.1111/ane.12368.CrossRefGoogle Scholar
  47. O’Brien, K. K., Schuttke, A., Alhakeem, A., Donnelly-Swift, E., Keogh, C., O’Carroll, A., et al. (2015). Health, perceived quality of life and health services use among homeless illicit drug users. Drug and Alcohol Dependence, 154, 139–145.  https://doi.org/10.1016/j.drugalcdep.2015.06.033.CrossRefGoogle Scholar
  48. Padden, D. L., Connors, R. A., Posey, S. M., Ricciardi, R., & Agazio, J. G. (2013). Factors influencing a health promoting lifestyle in spouses of active duty military. Health Care for Women International, 34, 674–693.  https://doi.org/10.1080/07399332.2012.736572.CrossRefGoogle Scholar
  49. Parker, J., Reitzes, D. C., & Ruel, E. E. (2016). Preserving and protecting well-being among homeless men. Sociological Perspectives, 59, 201–218.  https://doi.org/10.1177/0731121415591096.CrossRefGoogle Scholar
  50. Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19, 2–21.CrossRefGoogle Scholar
  51. Peter, C., Cieza, A., & Geyh, S. (2014). Rasch analysis of the general self-efficacy scale in spinal cord injury. Journal of Health Psychology, 19, 544–555.  https://doi.org/10.1177/1359105313475897.CrossRefGoogle Scholar
  52. Pollard, E. L., & Davidson, L. D. (2001). Foundations of child well-being. UNESCO action research in family and early childhood series. Paris: UNESCO.Google Scholar
  53. Ponizovsky, A. M., Levov, K., Schultz, Y., & Radomislensky, I. (2011). Attachment insecurity and psychological resources associated with adjustment disorders. American Journal of Orthopsychiatry, 81, 265–276.  https://doi.org/10.1111/j.1939-0025.2011.01095.x.CrossRefGoogle Scholar
  54. Revelle, W. (2017). Psych: Procedures for personality and psychological research. Evanston: Northwestern University Retrieved from http://CRAN.R-project.org/package=psych.Google Scholar
  55. Roos, S. M., Potgieter, J. C., & Temane, M. Q. (2013). Self-efficacy, collective efficacy and the psychological well-being of groups in transition. Journal of Psychology in Africa, 23, 561–567.CrossRefGoogle Scholar
  56. Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48, 1–36.CrossRefGoogle Scholar
  57. Rowe, M., Styron, T., & David, D. H. (2016). Mental health outreach to persons who are homeless: Implications for practice from a statewide study. Community Mental Health Journal, 52, 56–65.  https://doi.org/10.1007/s10597-015-9963-4.CrossRefGoogle Scholar
  58. Ryff, C. D. (1989). Happiness is everything, or is it? Explorations on the meaning of psychological well-being. Journal of Personality and Social Psychology, 57, 1069–1081.CrossRefGoogle Scholar
  59. Ryff, C. D., & Keyes, C. L. M. (1995). The structure of psychological well-being revisited. Journal of Personality and Social Psychology, 69, 719–727.CrossRefGoogle Scholar
  60. Ryff, C. D., & Singer, B. (1996). Psychological well-being: Meaning, measurement, and implications for psychotherapy research. Psychotherapy and Psychosomatics, 65, 14–23.CrossRefGoogle Scholar
  61. Sadow, D., & Hopkins, B. (1993). Resiliency training and empowerment among homeless, substance-abusing veterans. Research Communications in Psychology, Psychiatry and Behaviour, 18, 121–134.Google Scholar
  62. Scherbaum, C. A., Cohen-Charash, Y., & Kern, M. J. (2006). Measuring general self-efficacy: A comparison of three measures using item response theory. Educational and Psychological Measurement, 66, 1047–1063.CrossRefGoogle Scholar
  63. Scholz, U., Gutiérrez-Doña, B., Sud, S., & Schwarzer, R. (2002). Is general self-efficacy a universal construct? Psychometric findings from 25 countries. European Journal of Psychological Assessment, 18, 242–251.CrossRefGoogle Scholar
  64. Schönfeld, P., Brailovskaia, J., Bieda, A., Zhang, X. C., & Margraf, J. (2016). The effects of daily stress on positive and negative mental health: Mediation through self-efficacy. International Journal of Clinical and Health Psychology, 16, 1–10.  https://doi.org/10.1016/j.ijchp.2015.08.005.CrossRefGoogle Scholar
  65. Schwarzer, R., & Born, A. (1997). Optimistic self-beliefs: Assessment of general perceived self-efficacy in thirteen cultures. World Psychology, 3, 177–190.Google Scholar
  66. Schwarzer, R., & Jerusalem, M. (1995). Generalized self-efficacy scale. In J. Weinman, S. Wright, & M. Johnston (Eds.), Measures in health psychology: A user’s portfolio (pp. 35–37). Windsor: NFER-NELSON.Google Scholar
  67. Schwarzer, R., & McAuley, E. (2016). The world is confounded: A comment on Williams and Rhodes (2016). Health Psychology Review, 10, 133–135.  https://doi.org/10.1080/17437199.2016.1162667.CrossRefGoogle Scholar
  68. Schwarzer, R., Mueller, J., & Greenglass, E. (1999). Assessment of perceived general self-efficacy on the Internet: Data collection in cyberspace. Anxiety, Stress & Coping: An International Journal, 12, 145–161.  https://doi.org/10.1080/10615809908248327.
  69. Seilheimer, T. A., & Doyal, G. T. (1996). Self-efficacy and consumer satisfaction with housing. Community Mental Health Journal, 32, 549–559.  https://doi.org/10.1007/BF02251065.CrossRefGoogle Scholar
  70. Shakespeare-Finch, J., Rees, A., & Armstrong, D. (2015). Social support, self-efficacy, trauma and well-being in emergency medical dispatchers. Social Indicators Research, 123, 549–565.  https://doi.org/10.1007/s11205-014-0749-9.CrossRefGoogle Scholar
  71. Shear, B. R., & Zumbo, B. D. (2013). False positives in multiple regression: Unanticipated consequences of measurement error in the predictor variables. Educational and Psychological Measurement, 73, 733–756.  https://doi.org/10.1177/0013164413487738.CrossRefGoogle Scholar
  72. Sheer, V. C. (2014). A meta-synthesis of health-related self-efficacy instrumentation: Problems and suggestions. Journal of Nursing Measurement, 22, 77–93.CrossRefGoogle Scholar
  73. Sherer, M., Maddux, J. E., Mercandante, B., Prentice-Dunn, S., Jacobs, B., & Rogers, R. W. (1982). The self-efficacy scale: Construction and validation. Psychological Reports, 51, 663–671.  https://doi.org/10.2466/pr0.1982.51.2.663.CrossRefGoogle Scholar
  74. Slocum-Gori, S. L., & Zumbo, B. D. (2011). Assessing the unidimensionality of psychological scales: Using multiple criteria from factor analysis. Social Indicators Research, 102, 443–461.  https://doi.org/10.1007/s11205-010-9682-8.CrossRefGoogle Scholar
  75. Slocum-Gori, S. L., Zumbo, B. D., Michalos, A. C., & Diener, E. (2009). A note on the dimensionality of quality of life scales: An illustration with the satisfaction with life scale (SWLS). Social Indicators Research, 92, 489–496.  https://doi.org/10.1007/s11205-008-9303-y.CrossRefGoogle Scholar
  76. Smith, M. S., Wallston, K. A., & Smith, C. A. (1995). The development and validation of the perceived health competence scale. Health Education Research, 10, 51–64.  https://doi.org/10.1093/her/10.1.51.CrossRefGoogle Scholar
  77. Steenkamp, J.-B. E. M., & Baumgartner, H. (1998). Assessing measurement invariance in cross-national consumer research. Journal of Consumer Research, 25, 78–90.CrossRefGoogle Scholar
  78. Su, Z., Khoshnood, K., & Forster, S. H. (2015). Assessing impact of community health nurses on improving primary care use by homeless/marginally housed persons. Journal of Community Health Nursing, 32, 161–169.  https://doi.org/10.1080/07370016.2015.1057082.CrossRefGoogle Scholar
  79. van Seggelen-Damen, I., & van Dam, K. (2016). Self-reflection as a mediator between self efficacy and well-being. Journal of Managerial Psychology, 31, 18–33.CrossRefGoogle Scholar
  80. Walter, Z. C., Jetten, J., Parsell, C., & Dingle, G. A. (2015). The impact of self-categorizing as 'homeless' on well-being and service use. Analyses of Social Issues and Public Policy (ASAP), 15, 333–356.  https://doi.org/10.1111/asap.12089.CrossRefGoogle Scholar
  81. Washington, O. G. M., & Moxley, D. P. (2003). Group interventions with low-income African American women recovering from chemical dependency. Health and Social Work, 28, 146–156.CrossRefGoogle Scholar
  82. Washington, O. G. M., & Moxley, D. P. (2013). Self-efficacy as a unifying construct in nursing-social work collaboration with vulnerable populations. Nursing Inquiry, 20, 42–50.CrossRefGoogle Scholar
  83. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070.CrossRefGoogle Scholar
  84. Williams, D. M., & Rhodes, R. E. (2016). The confounded self-efficacy construct: Conceptual analysis and recommendations for future research. Health Psychology Review, 10, 113–128.  https://doi.org/10.1080/17437199.17432014.17941998.CrossRefGoogle Scholar
  85. Wu, C. (2009). Factor analysis of the general self-efficacy scale and its relationship with individualism/collectivism among twenty-five countries: Application of multilevel confirmatory factory analysis. Personality and Individual Differences, 46, 699–703.  https://doi.org/10.1016/j.paid.2009.01.025.CrossRefGoogle Scholar
  86. Zimmerman, B. (2000). Self-efficacy: An essential motive to learn. Contemporary Educational Psychology, 25, 82–91.  https://doi.org/10.1006/ceps.1999.1016.CrossRefGoogle Scholar
  87. Zimmerman, D. W., & Zumbo, B. D. (2015). Resolving the issue of how reliability is related to statistical power: Adhering to mathematical definitions. Journal of Modern Applied Statistical Methods, 14, 9–26.CrossRefGoogle Scholar
  88. Zumbo, B. D. (2007). Validity: Foundational issues and statistical methodology. In C. R. Rao & S. Sinharay (Eds.), Handbook of statistics (Vol. 26: Psychometrics, pp. 45–79). Amsterdam: Elsevier Science B.V.Google Scholar
  89. Zumbo, B. D. (2009). Validity as contextualized and pragmatic explanation, and its implications for validation practice. In R. W. Lissitz (Ed.), The concept of validity: Revisions, new directions and applications (pp. 65–82). Charlotte: IAP – Information Age Publishing, Inc..Google Scholar
  90. Zumbo, B. D., & Hubley, A. M. (2016). Bringing consequences and side effects of testing and assessment to the foreground. Special issue on Validity. Assessment in Education: Principles, Policy, & Practice, 23, 299–303.  https://doi.org/10.1080/0969594X.2016.1141169.
  91. Zumbo, B. D., & Hubley, A. M. (Eds.). (2017). Understanding and investigating response processes in validation research. New York: Springer.Google Scholar
  92. Zumbo, B. D., & Taylor, S. V. (1993). The construct validity of the extraversion subscales of the Myers-Briggs type indicator. Canadian Journal of Behavioural Science, 25, 590–604.  https://doi.org/10.1037/h0078847.CrossRefGoogle Scholar
  93. Zumbo, B. D., Gelin, M. N., & Hubley, A. M. (2002). The construction and use of psychological tests and measures. Encyclopedia of life support systems (EOLSS): Psychology theme (pp. 1–28). Oxford: Eolss Publishers. http://www.eolss.net [entry,6.27.3.4].
  94. Zumbo, B. D., Gadermann, A. M., & Zeisser, C. (2007). Ordinal versions of coefficients alpha and theta for likert rating scales. Journal of Modern Applied Statistical Methods, 6, 21–29.  https://doi.org/10.1107/S0907444909031205.CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of ECPSUniversity of British ColumbiaVancouverCanada

Personalised recommendations