ABIEBR :: 17.19 Satisfaction with Life Scale (SWLS)

17.19 Satisfaction with Life Scale (SWLS)

Life satisfaction may be defined as a conscious, cognitive, global judgement of one’s own life.  It is not an assessment based on externally imposed objective standards, but rather depends upon a comparison of one’s life circumstances to one’s own internal standards or criteria (Diener et al. 1985, Pavot et al. 1991, Pavot and Diener 1993).  The Satisfaction with Life Scale was created to assess a person’s global judgment of life satisfaction (Diener et al. 1985). 

Diener et al. (1985) generated 48 self-report items related to satisfaction with life, including items assessing positive and negative affect.  Factor analyses were used to identify 3 factors; life satisfaction, negative affect and positive affect.  All affect items were eliminated as were items with factor loadings of less than 0.60.  The remaining 10 items were reduced to 5 on the basis of “semantic similarity” (Diener et al. 1985). 

Table 17.45  Satisfaction with Life Scale*

  1. In most ways my life is close to my ideal
  2. The conditions of my life are excellent.
  3. I am satisfied with my life. 
  4. So far I have gotten the important things I want in life. 
  5. If I could live my life over, I would change almost nothing. 

  *(Diener et al. 1985)

Respondents are instructed to rate each item using a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree).  Item ratings are summed to provide a total score ranging from 5 – 35 where higher scores are indicative of greater life satisfaction.  The SWLS takes a global approach to assessment.  Because no specific domains are named within the scale and items are not specific in nature, the respondent remains free to consider the life domains or affective components he or she feels make the most important contribution to their subjective experience of happiness (Diener et al. 1985, Pavot and Diener 1993, Arrindell et al. 1999). 

The scale is short and simple to administer and score.  It can easily be added to assessments using multiple measures with no significant increase in time (Pavot et al. 1991). 

Table 17.46 Characteristics of the Satisfaction with Life Scale

Reliability

 

 

  • Test-retest: 0.82 for a 2-month interval (Diener et al. 1985); 0.84 for a 2-week interval and 0.84 for a one-month interval (student sample - Pavot et al. 1991); In their 1993 review, Pavot & Diener (1993) reported test retest reliability ranging 0.83 – 0.50 – intervals ranged from 2 weeks to 4 years and, in general, higher reliabilities were associated with shorter retest intervals. 
  • Internal consistency: Item to total correlations ranged from 0.57 – 0.75 (α=0.87) in a sample of undergraduate university students and from 0.63 – 0.81 in a sample of elderly persons (Diener et al. 1985); In a sample of older individuals (mean age = 74), α  = 0.83 and 0.85 in a sample of university students (Pavot et al. 1991); Item to total correlations ranged from 0.55 – 0.80 among older individuals and 0.63 – 0.77 among the students; α=0.91(time 1) and 0.82(time 2) – points separated by a few weeks (Suh & Diener 1996); reliability according to Fleishman & Benson formula (1987) was 0.921 (Shevlin et al. 1998); Arrindell et al. (1999) reported α = 0.82 and item-total correlations ranging from 0.5 to 0.7;  α=0.78 for the Portuguese version in an adolescent sample (Neto et al. 1993); In a review, Pavot & Diener identified 6 articles evaluating internal consistency – α ranged from 0.79 – 0.89;  item-to-total correlations ranged from 0.71 to 0.86—mean interitem correlation = 0.70 a= 0.92 (Westaway et al. 2003);  Lucas et al. (1996) reported a = 0.84, 0.84 and 0.88 over three studies; a = 0.78, mean inter item correlation = 0.41, item total correlations ranging from 0.52 – 0.65 (Neto 1993); a=0.86 (Meyer et al. 2004).

Validity

  • Construct validity: Principal components factor analysis (PCA) revealed a single factor accounting for 66% of the variance – factor loadings ranged from 0.61(item 5) to 0.84 (Item 1) (Diener et al. 1985); PCA revealed a single factor accounting for 65% and 74% of variance in elderly and student subject samples respectively – loadings ranged from 0.78 – 0.93 (Pavot et al. 1991); PCA revealed a single factor accounting for 60.1% of variance – items 1-4 factor loadings > 70%, item 5 = 0.64 (Arrindell et al. 1999); factor analysis revealed a single factor accounting for 76% of the variance – factor loadings ranged from 0.81 to 0.92 (Westaway et al. 2003); a one-factor measurement model was found for both male and female Spanish adolescents suggesting no factor invariance across the sexes (Atienza et al. 2003); Shevlin et al. (1998) reported a single factor with factor loadings ranging from 0.92 to 0.96; PCA analysis revealed a single factor accounting for 53.3% of variance (Neto 1993);  
  • Construct validity (known groups): SWLS scores differentiated between groups of young adults defined by marital status (p<0.001) (Arrindell et al. 1999); significant differences in life satisfaction were identified between all groups of patients based on analyzed disorder (substance use, affective disorder, anxiety disorder, somatoform disorder) and those with no disorder (Meyer et al. 2004).
  • Construct validity (convergent/divergent): SWLS scores correlated with selected personality measures: 0.54 with self-esteem, -0.41 with symptom checklist, -0.48 with neuroticism, -0.25 with emotionality, 0.20 with sociability and very low correlations with activity and impulsivity (Diener et al. 1985); r=0.86 with rated self-esteem (Westaway et al. 2003); r=0.52 (time 1), 0.43 (time 2) with positive affect and r=-0.36 (time 1), -0.30 (time 2) with negative affect  (Lucas et al. 1996); r=0.60 (time 1) and 0.52 (time 2) with optimism and r=0.59 (time 1) and 0.55 (time 2) (Lucas et al. 1996); SWLS scores correlated with global happiness (Fordyce Scale r=0.68) as well as with affect balance (r=0.76; Pavot et al. 1991); multi-method multi-trait analyses demonstrated that assessment via the SWLA is able to discriminate between life satisfaction and both affective aspects of SWB, optimism and self-esteem (Lucas et al. 1996); significant positive correlations were demonstrated with social acceptance, self-efficacy, psychological maturity, impulsivity/activity, self-concept, physical attractiveness and happiness while significant negative correlations between SWLS and loneliness, self-assessed loneliness, social anxiety and shyness were reported (Neto 1999); SWLS scores correlated with recent (within 3 months) positive and negative life events (r=0.25 and –0.28, respectively, p<0.01) (Suh et al. 1996).
  • Concurrent validity: Moderately strong correlations (r=0.47-0.68) with other measures of subjective well-being, including: Fordyce’s % of time happy question and single-item measure of happiness, Differential Personality Questionnaire, Cantril’s Self-Anchoring Ladder, Gurin, Andrews and Withey’s D-T scale, Campbell, Bradburn’s Affect Balance Scale, and Summed Domain Satisfaction – in addition, SWLS scores correlated with interviewer rating of life satisfaction (r=0.43) (Diener et al. 1985); Pavot et al. (1991) reported moderate to strong correlations (r=0.42 – 0.81) with both self and peer reported assessments of life satisfaction (LSI-A, Philadelphia Geriatric Morale Scale, Daily satisfaction, memory difference, peer-rated SWLS & peer-rated LSI-A ); In a review of studies evaluating SWLS, Pavot & Diener (1993) reported convergence with related measures (Andrews/Withey Scale, Fordyce Global Scale) as well as negative correlations with measures of distress (Beck Depression Inventory, negative affect and anxiety, depression & distress on the Symptom Checklist-90); r=0.56 with Oxford Happiness Inventory, r=0.61 with the Depression-Happiness scale and neuroticism and conscientiousness were the most significant predictors of SWLS scores (Hayes et al. 2003).

Responsiveness

- from beginning of therapy to one month into the therapy process, SWLS scores changed significantly for clients (p<0.01, n=7)(Friedman, 1991 in Pavot & Diener 1991); elderly caregivers of patients with dementia demonstrated significant decline in satisfaction with life scores over time (p<0.05) (Vitaliano et al.1991).

Tested for TBI patients?

No

Other Formats

The Extended Satisfaction with Life Scale (ESWLS)(Alfonso et al. 1996, Gregg and Salisbury 2001)
The Temporal Satisfaction with Life Scale (TSWLS) (Pavot et al. 1998, McIntosh 2000)

Use by proxy?

Pavot et al. (1991) reported correlations between self and peer rated SWLS scores (r=0.54) when used to assess elderly individuals (mean age = 74).  Among a student population, correlation between peer reports and family reports = 0.54, between self report and peer report = 0.55 and between self and family report = 0.57. 

Advantages.  The scale is available freely and is simple to administer and score.  With only 5 items, it takes very little time to complete.  The scale has been evaluated for use in populations of varying ages (adolescent, young adult and senior).  Scale items appear to be at the 6th – 10th grade reading level, making it comprehensible to a wide range of adults.  Original scale was tested in both college student and geriatric populations (Diener et al. 1985).  Scale items are at the 6th to 10th grade reading level, which makes it comprehensible to most adults (Pavot and Diener 1993).  The scale has been evaluated in several cultures and has been translated into several languages including Dutch, Taiwanese, Spanish, French, Russian, Korean, Hebrew, Mandarin Chinese, Spanish, Spanish and Portuguese. 

It has been suggested that social desirability may account for a large port of variance in the assessment of subjective well-being and may, in fact be an important component of well-being (Pavot and Diener 1993).  However, Diener et al. (1985) reported a very weak association between SWLS scores and the Marlowe-Crowne scale of social desirability (r=0.02). 

Limitations.  Although the SWLS is used to evaluate satisfaction with life in populations of adults with acquired brain injury, no studies specifically evaluating the use of this scale within this specific population could be identified. 

While the SWLS is a simple scale, interpretation of scores is not clear.  The SWLS was not intended to provide an assessment of subjective well-being (SWB), only a single aspect of well-being.  One cannot assume that SWLS scores provide a direct assessment of emotional well-being.  In order to assess the broader construct of SWB, assessment of negative and positive affect should be included (Pavot and Diener 1993).  Furthermore, no published normative data for the SWLS could be located.  Pavot and Diener (1993) identified numerous studies providing means and standard deviations for SWLS scores in a variety of populations and note considerable variation within different population subsets.  However, scores may be interpreted in absolute rather than relative terms.  In this case, it has been suggested that a score of 20 is regarded as neutral, while scores in excess of 20 represent satisfaction (21-25=slightly satisfied; 26-30= satisfied), and scores of less than 20 represent dissatisfaction (15-19=slightly dissatisfied; 5-9=extremely dissatisfied) (Pavot & Diener 1993). 

The SWLS does not appear to be affected by gender or age (Pavot and Diener 1993).  Factor analyses focusing on factorial invariance across gender has demonstrated that the structure and measurement of life satisfaction are equivalent across groups, that is, the strength of relationships between items and the underlying construct is the same for men and women (Wu and Yao 2006, Shevlin et al. 1998).  However, factorial invariance was not demonstrated on evaluation of the Spanish version of the SWLS (Atienza et al. 2003, Pons et al. 2000).  Westaway et al. (2003) reported that SWLS scores were not related to either gender or age, but rather to employment status and level of education.  Similarly, Neto (1993) identified significant main effects associated with both gender and socioeconomic status (SES) such that higher SES and male gender were associated with greater satisfaction with life as assessed on the SWLS. 

Summary – Satisfaction with Life Scale

Practicality
Interpretability.  Guidelines for absolute interpretation of scores are available.  To our knowledge, no normative data is presently available for the SWLS. 
Acceptability.  Scale items are at a suitable reading level for most adults and it takes a minimal amount of time for the subject to complete the measure in its entirety.
Feasibility.  Brief, simple, and low-cost administration.

Table 17.47  SWLS Evaluation Summary
 


Reliability

Validity

Responsiveness

Rigor

Results

Rigor

Results

Rigor

Results

Floor/ceiling

+++

 

++ (TR)
+++ (IC)

++

+++

+

+

n/a

NOTE: +++=Excellent; ++=Adequate; +=Poor; n/a = insufficient information; TR=Test re-test; IC= internal consistency; IO = Interobserver;