Hypothesis: A multidimensional measure of wellbeing will consist of 8 equally-weighted, contributing dimensions; there will be shared covariance for the dimension dyads:

1. Psychological & Intellectual Wellbeing with Mental & Emotional Wellbeing

2. Mental & Emotional Wellbeing with Spiritual & Holistic Wellbeing

3. Life Satisfaction & Meaning with Spiritual & Holistic Wellbeing

4. Occupational Wellbeing with Financial Wellbeing

Proposed Dimensions: The following 8 dimensions were identified as the foundation for a comprehensive, multidimensional measurement of wellbeing, as well as what defines them:

1. Psychological & Intellectual Wellbeing – relating to the cognitive processes associated with one’s mental and emotional state

2. Physical Wellness – relating to physical health

3. Mental Health & Emotional Wellbeing – relating to the feelings and self-perceptions of one’s mental and emotional health

4. Life Satisfaction & Meaning – relating to how one evaluates their life as a whole, rather than acute emotional state

5. Spiritual & Holistic Wellbeing – relating to one’s religious, spiritual, or holistic whole-body spirituality or feeling a part of something larger than oneself

6. Occupational Wellbeing – relating to wellbeing associated with satisfaction with one’s occupation and the relationships with peers, inferiors, and superiors

7. Financial Wellbeing – relating to one’s self-evaluation of socioeconomic status and financial freedom or dependence

8. Social & Environmental Wellbeing – relating to evaluation of one’s place in the world and how they interconnect with the environment and society

Measures: In this initial phase, all items from all scales were used. The following scales were utilized for the purpose of creating a question bank that the dimensions identified would draw from (click here to view complete scales):

1. Basic Needs Satisfaction in General Scale (Ryan & Deci, 2000) – 21-item measure that was created to assess the satisfaction of basic psychological needs in general. Participants are instructed to indicate how true they felt each statement was of their life and respond on a scale of 1 (Not at all true) to 7 (Very true). Nine of the 21 items are negatively worded and were reversed scored prior to analyses. Higher scores are indicative of a higher level of satisfaction of needs. Reliability of the measure ranged from Cronbach’s α = .55 to .82 for the three subscales in three distinct samples, though information regarding the construct validity is mixed in the literature. In the original study, only 3 of 7 external measures correlated significantly to support the hypotheses. Permission for use and modification granted 25 September 2023.

2. PERMA Profiler (Butler & Kern, 2016) – 23-item measure created to assess five pillars of wellbeing (positive emotion, engagement, relationships, meaning, accomplishment). Questions are formatted on an 11-point scale with polar end labels (e.g., 0 indicating “not at all”, 10 indicating “completely”). Reliability of the measure is high with α = .94 across pilot studies (N = 31,966), and construct validity is satisfied with correlations ranging from .80 to .84 for both convergent and discriminant measures. Permission for use and modification granted 26 September 2023.

3. Mind-Body-Spirit Wellness Behavior & Characteristics Index (Hey et al., 2006) – 44-item measure using a mind-body-spirit framework to assess behaviors and characteristics associated with wellbeing. Participants are instructed to indicate their agreement with statements on a 3-point Likert scale with 1 indicating “rarely/seldomly”, 2 indicating “occasionally/sometimes” and 3 indicating “often/always”. Reliability and construct validity of the measure are acceptable, with α = .81 - .91, and correlations with convergent and discriminant measures ranging from .73 to .86. Permission for use and modification granted 26 September 2023.

4. Professional Quality of Life (Stamm, 2009) – 30-item measure developed in order to examine quality of life regarding compassion satisfaction and compassion fatigue in the helping professions. Questions are formatted on a Likert scale from 1 (indicating “never”) to 5 (indicating “very often”), and participants are asked to consider the questions in the context of their current work situations”. Reliability and construct validity of the measure are acceptable, with α = .80 to .90 across subscales, and correlations with convergent and discriminant measures ranging from .43 to .69. Permission for use and modification granted 20 June 2024.

5. Fulfilled Life Scale (Baumann & Ruch, 2022) – 32-item measure that was created to explore an individual’s self-perceptions about their lives lived so far. The first 24 questions ask participants to refer to their lives lived in retrospect, and the final 8 questions ask about their feelings about looking back on their life. Responses are recorded on a 6-item Likert scale with 1 indicating “does not apply at all” and 6 indicating “applies completely”. Reliability and construct validity of the measure are acceptable, with α = .82 to .92, and correlations with convergent and discriminant measures ranging from .28 to .77. Permission for use and modification granted 29 September 2023.

6. Warr’s Job Satisfaction Scale (Warr et al., 1979) – 16-item measure designed to examine an individual’s satisfaction with different aspects of their occupation. Responses are recorded on a 7-item Likert scale with 1 indicating “extremely dissatisfied" and 7 indicating “extremely satisfied”. Reliability and construct validity of the measure are acceptable, with α = .80 to .91, and significant correlations with convergent and discriminant measures, though specific correlations were not reported. Permission for use and modification requested 25 September 2023.

7. Workplace Belonging Scale (Jena & Pradham, 2018) – 12-item measure created with the intent of having a context-specific sense of belonging scale specific to an individual’s occupation and workplace setting. All items were positively worded, and participants were asked to response on a Likert-style scale of 1, indicating “strongly disagree” to 5, indicating “strongly agree”. Reliability and construct validity of the measure are acceptable, with α = 0.86, and correlations with convergent and discriminant measures of at least r = .38. Permission for use and modification granted 25 September 2023.

8. Financial Wellbeing Scale (Consumer Financial Protection Bureau, 2017) – 10-item measure aimed at examining financial wellbeing in individuals, described as the ability for an individual to fully meet their current and ongoing financial obligations. Items are split into two components. For the first 6 questions, participants were asked to rate how well statements described their experiences on a Likert-style scale of 1, indicating “describes me completely”, to 6, indicating “does not describe me at all”. The final 4 questions asked participants to rate how often a statement applied to them, with 1 indicating “always” and 6 indicating “never”. Reliability and construct validity of the measure are acceptable, with α = .89, and correlations with convergent and discriminant measures ranging from .30 to .64. Use of scale is granted open access via government institution; and additional permission for modification was granted 16 October 2023.

9. Social Wellbeing Scale (Keyes, 1998) – 33-item measure intended to examine the impact of social wellbeing, social integration, social contribution, social coherence, social actualization, and social acceptance. Participants were asked to rate how much they agreed or disagreed with the statements. Responses were recorded on a Likert-style scale of 1, indicating “strongly disagree” to 7, indicating “strongly agree”. . Reliability of the measure is borderline, with α = 0.57 in the first study, and α = 0.64 in the second study, and construct validity is acceptable correlations with convergent and discriminant measures of at least r = 0.30. Permission for use and modification granted 1 October 2023.

Software: IBM's Statistical Package for the Social Sciences (SPSS; version 27)

Statistical Methods: Initially, there were a total of 180 items in their original form. There were a total of 5 rounds of item analysis, with the goal of reaching 100 items for the total measure. Whether items were kept or removed was directly tied to the literature and theoretical framework of the wellbeing construct. Item removal was conducted in two parts; the first component consisted of 4 rounds of theory-driven removal, and the second component comprised of 1 round of reliability analysis repeated until reliability was satisfactory for the combined scale as well as the dimension subscales. The selection criteria for each round was as follows:

1. Theory-Driven Removal of Items:

   1. 1. Round 1 - All items from the 8 contributing measures were listed by dimension. Each item was discussed among the research team, and items that best described the dimension they were part of were highlighted as items to keep. The purpose of this round was not to remove items, but to identify the items that were the most descriptive of the dimension being assessed. At the end of Round 1: N  = 180 total scale items

      2. Round 2 - The purpose of the 2nd round was to remove items that did not capture the essence of the dimension they represented; 5 items were removed from psychological/intellectual, 4 items from physical, 12 items from mental/emotional, 5 from life satisfaction/meaning, 6 from spiritual/holistic, 4 from occupational, and 0 from social/environmental.  At the end of Round 2: N  = 144 total scale items

      3. Round 3 - Similar to round one, round three was intended as a way to solidify the items that best described each dimension. Each item was discussed in detail again, and those best describing their dimensions were highlighted to keep. Items not highlighted after this round were removed: 7 from psychological/intellectual, 2 from physical, 14 from mental/emotional,  12 from life satisfaction/meaning, 2 from spiritual/holistic, 8 from occupational, 0 from financial, and 8 from social/environmental. The Professional Quality of Life Scale (30 items; Stamm, 2009) and the Workplace Belonging Scale (12 items; Jena & Pradham, 2018) were added at this time, with the intent of analyzing them in round 4. At the end of Round 3: N  = 106 + 42 (new) = 148 total scale items

      4. Round 4 - In the final round of theory-driven removal, the purpose shifted back to determining if each dimension subscale was being captured by the items kept during previous rounds. Novel questions were added at this time. At this time, items that needed rewording or adjustment were also flagged, but no further items were removed. At the end of Round 4: N  = 116 total scale items

2. Reliability Analysis:

1. Inter-item correlations were conducted first in order to reduce items prior to factor analysis, which removed 6 items that were correlated with other items at r ≥ .800.

2. After correlational findings, reliability analysis was conducted independently on each subscale using the following steps:

   1. 1. Reliability analysis was conducted on the complete in order to determine the baseline Cronbach’s α for that particular subscale

      2. The item indicating the highest change in α if item removed was selected for removal from the subscale

      3. Once the item was removed, reliability analysis of the subscale was conducted again without the removed item

      4. This process was repeated until there were no longer any items that were detrimental to the subscales’ α, and considered complete once no further items were necessary for removal; ten items total were removed from the subscales utilizing this method.

Once complete, reliability assessment was conducted for each subscale independently, and data from the measures were found to have acceptably reliability, reflecting Cronbach’s α of at minimum of .771, and majority of subscales reflecting an α > .900. Finally, reliability analysis was conducted for the full measure of multidimensional wellbeing, and results indicated an overall Cronbach’s α of .962 (α∆ = .001), suggesting that the removal of items from the subscales was not detrimental to the overall measure.  

At the completion of of Reliability Analysis: N  = 99 total scale items

Scaling: A 6-point Likert style scale was used for responses (1 = strongly disagree, 2 = disagree, 3 = somewhat disagree, 4 = somewhat agree, 5 = agree, 6 = strongly agree), and all items were adjusted or reworded so that they could utilize the same response scale. Items that may be candidates for removal were highlighted so that attention could be brought to them after factor analysis and item correlations were conducted. Permission was obtained for the purpose of modifying each of the scales used for this purpose.