Free
Research Article
Issue Date: April 08, 2015
Published Online: April 13, 2015
Updated: January 01, 2020
Development and Evaluation of the Lifestyle History Questionnaire (LHQ) for People Entering Treatment for Substance Addictions
Author Affiliations
  • Linda M. Martin, PhD, OTR, was Professor and Chair, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study and is now retired; lmartin@fgcu.edu
  • Robert Triscari, PhD, is Assistant Professor, College of Education, Florida Gulf Coast University, Ft. Myers
  • Rosemary Boisvert, MS, OTR, CAP, is Assistant Vice President for Residential Services, SalusCare, Inc. (formerly Southwest Florida Addiction Services), Ft. Myers, FL
  • Kristi Hipp, MSOT; Jennifer Gersten, MSOT; Rachel C. West, MSOT; Elizabeth Kisling, MSOT; Aaron Donham, MSOT; Naomi Kollar, MSOT; and Patricia Escobar, MSOT, were students, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study
  • Kristi Hipp, MSOT; Jennifer Gersten, MSOT; Rachel C. West, MSOT; Elizabeth Kisling, MSOT; Aaron Donham, MSOT; Naomi Kollar, MSOT; and Patricia Escobar, MSOT, were students, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study
  • Kristi Hipp, MSOT; Jennifer Gersten, MSOT; Rachel C. West, MSOT; Elizabeth Kisling, MSOT; Aaron Donham, MSOT; Naomi Kollar, MSOT; and Patricia Escobar, MSOT, were students, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study
  • Kristi Hipp, MSOT; Jennifer Gersten, MSOT; Rachel C. West, MSOT; Elizabeth Kisling, MSOT; Aaron Donham, MSOT; Naomi Kollar, MSOT; and Patricia Escobar, MSOT, were students, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study
  • Kristi Hipp, MSOT; Jennifer Gersten, MSOT; Rachel C. West, MSOT; Elizabeth Kisling, MSOT; Aaron Donham, MSOT; Naomi Kollar, MSOT; and Patricia Escobar, MSOT, were students, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study
  • Kristi Hipp, MSOT; Jennifer Gersten, MSOT; Rachel C. West, MSOT; Elizabeth Kisling, MSOT; Aaron Donham, MSOT; Naomi Kollar, MSOT; and Patricia Escobar, MSOT, were students, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study
  • Kristi Hipp, MSOT; Jennifer Gersten, MSOT; Rachel C. West, MSOT; Elizabeth Kisling, MSOT; Aaron Donham, MSOT; Naomi Kollar, MSOT; and Patricia Escobar, MSOT, were students, Department of Occupational Therapy and Community Health, Florida Gulf Coast University, Fort Myers, at the time of the study
Article Information
Mental Health / Mental Health
Research Article   |   April 08, 2015
Development and Evaluation of the Lifestyle History Questionnaire (LHQ) for People Entering Treatment for Substance Addictions
American Journal of Occupational Therapy, April 2015, Vol. 69, 6903250010. https://doi.org/10.5014/ajot.2015.014050
American Journal of Occupational Therapy, April 2015, Vol. 69, 6903250010. https://doi.org/10.5014/ajot.2015.014050
Abstract

OBJECTIVE. We developed and investigated the psychometric properties of the Lifestyle History Questionnaire (LHQ), a self-report instrument designed to measure the extent of occupational dysfunction attributable to substance abuse.

METHOD. The instrument was developed using concepts in the ecological models of occupational therapy and in the work of William L. White, who defined addiction culture in terms of the patterns of life in context. We analyzed data from two field tests using both classical test theory and item response theory.

RESULTS. The final version of the instrument has 70 items, 1 unifying construct, and 8 subscales. We found it to be valid and reliable (α = .93) for measuring the extent of occupational dysfunction and specific areas of strengths and weaknesses.

CONCLUSION. The LHQ is a promising new instrument, the first of its kind to measure occupational dysfunction in context for people with substance addictions.

William L. White (1996), a well-known author in the field of addiction research, described the culture of addiction as “an informal social network in which group norms (prescribed patterns of perceiving, thinking, feeling, and behaving) promote excessive drug use” (p. 5); people who become addicted are “enmeshed” in an addiction culture. White proposed an assessment and treatment model based on multiple pathways of addiction etiology. The model generalizes addiction culture into three etiological pathways, each of which is further divided into subcultures: (1) The acultural addict, “a person who has initiated and sustained addiction in isolation from other addicts” (p. 10); (2) the culturally enmeshed addict, a person who is immersed in a social group that does not ascribe to mainstream social values, in which “abstinence or moderation are the prevailing attitudes regarding psychoactive substances” (p. 12); and (3) the bicultural addict, a person who exists in both “mainstream culture and the culture of addiction” (p. 14), depending on the physical and social environment at the time.
White (1996)  further described the necessity for a “culture of recovery” to be established and populated by those who are in treatment for addiction to address the extensive contextual and behavioral aspects necessary for successful recovery. This culture of recovery focuses on contextual factors in clients’ lives to a greater extent than traditional treatment. In White’s (2009)  words, “Mainstream services seek to modify the physiology, thoughts, feelings, and behaviors of individual service consumers with little effort extended to ‘treat’ the larger physical and relational worlds in which individual recovery efforts succeed or fail” (p. 146). White (2009)  emphasized the need for “sustained recovery maintenance in natural community environments” (p. 146).
Recent research has demonstrated that environment is perhaps the biggest factor in changing established habits (Neal, Wood, & Quinn, 2006; Wood, Tam, & Witt, 2005). The Occupational Therapy Practice Framework: Domain and Process (3rd ed.; American Occupational Therapy Association [AOTA], 2014) reflects a model of human functioning in all areas of daily life occurring in context (cultural, personal, temporal, and virtual) and within the environment (physical and social). Occupational therapy practice likewise is illustrated as occurring in context and as involving delivery of interventions that will help clients perform in the contexts and environments they encounter in their personal lives (AOTA, 2014).
Occupational therapy theories and models of practice have historically focused on functional behaviors in context to some degree, with increasing emphasis on environment and context. The Occupational Adaptation frame of reference (Schkade & Schultz, 1992) began the trend of increasing focus on environment and context; the ecological models Ecology of Human Performance (EHP; Dunn, Brown, & McGuigan, 1994), Person−Environment−Occupation (PEO; Law et al., 1996), and Person−Environment−Occupational Performance (PEOP; Christiansen & Baum, 1997) all expanded on that theme. These models resonate with White’s writings on a culture of recovery in that they reveal the transactional nature of the person and the environment but add occupational factors to the list of factors shaping occupational performance. A principal tenet of these models is that environment is a critical determinant of “successful and satisfying occupational performance” and that it can “either facilitate or inhibit occupational performance” (Brown, 2014, p. 499).
Both White (1996)  and Brown (2014)  advocated for assessment and adaptation of the environment as a primary intervention strategy to promote change in addictive behaviors and occupational performance, respectively. A necessary tool for evaluating the constructs White discussed and for measuring change through the process of recovery was needed. Although many instruments have been designed and used in addictions treatment (e.g., Addiction Severity Index, McLellan, Luborsky, Woody, & O’Brien, 1980) and several assessments have been designed to assess occupational performance in context (e.g., Occupational Performance History Interview, Kielhofner et al., 1998; Canadian Occupational Performance Measure, Law et al., 2005), these tools often require 45 min or more to complete, and no instrument developed thus far blends the concepts of the two professional perspectives.
The purpose of this study was to develop and test an instrument that could measure the degree of a person’s enmeshment in the culture of addiction or, from the perspective of an occupational therapy practitioner, the degree of occupational dysfunction in context attributable to substance abuse. The objective was to develop a relatively short self-report instrument that addresses occupational therapy domains of interest to assist practitioners in determining intervention strategies for their clients. This article reports on the development, testing, and refinement of the instrument.
Method
Research Design
This methodological study examined the psychometric properties of the Lifestyle History Questionnaire (LHQ). We obtained institutional review board approval before recruiting participants, and all participants provided signed consent to be included in the study. Principal component analysis, both exploratory and confirmatory factor analysis, and Rasch analysis were used to assess the structure of the instrument. We examined internal consistency (Cronbach’s α), content and procedural validity, construct validity, and item fit and dimensionality.
Internal consistency examines the extent to which the instrument has one underlying unidimensional trait being measured (Streiner, 2003). When a measure contains multiple subscores representing different factors, internal consistency is especially important because it reflects the degree to which the overall content is homogeneous (Lemke & Wiersma, 1976). Content and procedural validity were assessed by examining the process of development of the instrument itself. Construct validity was examined using factor analysis. Finally, Rasch analysis was used in Phase 2 of the study because of its specific appropriateness for evaluating item fit and dimensionality of rating scales.
Participants
We recruited adult participants from two facilities—a detoxification unit and a residential treatment center—treating people with substance addictions. We administered the instrument to groups of participants at the end of the detoxification program or in the 1st or 2nd wk of residential treatment to obtain data from people near the same stage of treatment and to capture a clear depiction of the occupational components of addiction before treatment programs produced significant changes and memory faded. Participation was voluntary and data were gathered anonymously; we emphasized these facts to participants and asked them to respond to the items honestly if they chose to participate.
Instrument
We developed a preliminary self-report instrument, the LHQ, to explore the culture of addiction as described by White (1996) . We used concepts from the EHP, PEO, and PEOP models because of their focus on the transactional nature of the tie between people and their performance of tasks and occupations in context. The instrument addresses the person (e.g., thoughts, feelings), the physical and social environment, and the cultural and temporal context, as well as patterns of behavior (tasks and occupational performance in context). All items are phrased as first-person statements—for example, “I felt like drugs/alcohol controlled my life” (person), “I preferred to hang out with people who were using drugs or alcohol” (social environment), and “My substance use contributed to problems at work” (tasks and occupational performance in context). The LHQ asks participants to rate the extent to which the statement reflects their experience within the recent past (1−2 mo) just prior to entering treatment.
A team of academic and clinical occupational therapists and occupational therapy students developed the items. The team shared the initial draft of the instrument with the professional staff of a large local addictions treatment agency to gain feedback regarding needed revisions. These professionals, who included occupational therapists, were well versed in White’s work, and all understood the focus on thoughts, behaviors, and environmental factors. Suggestions from the staff were incorporated into the instrument, and the final draft for Phase 1 testing comprised 127 items.
Development and testing of the LHQ was completed over the course of 3 yr in four stages: (1) development of items, (2) Phase 1 pilot testing, (3) refinement of the instrument, and (4) Phase 2 testing. During this time, the instrument was revised and edited using data from the two pilot administrations and expert judgment.
The current version of the instrument contains 70 items to be rated 4 (very often), 3 (frequently), 2 (once in a while), 1 (rarely), or 0 (never). The majority of items were phrased to reflect dysfunctional thoughts and behaviors often observed in people addicted to substances; for items that were stated to reflect residual strengths or positive thoughts and behaviors, we reversed the numeric scores so the directionality of the trait total score remained consistent. Therefore, high scores on the LHQ reflect deeper enmeshment in the culture of addiction (White, 1996) and severely disrupted occupational performance.
Phase 1 Data Collection, Analysis, and Revision
Phase 1 data collection gathered demographic variables, including gender, age, race and ethnicity, marital status, education level, preferred substance, and length of addiction, and responses to the 127-item version of the LHQ. The sample in Phase 1 field testing consisted of 232 people, of whom 3 returned incomplete questionnaires, for a final sample of 229. All participants were 18 yr and older, with the majority (65.1%) between ages 25 and 44 yr; 62.3% were male; and 83.1% were White, 5.8% Native American, 5.3% African-American, 4.4% Hispanic, and 1.3% other or mixed. The majority, 73.2%, reported having been dependent on drugs or alcohol 5 or more yr.
Data analysis was supported by PASW Statistics for Windows Version 18.0 (SPSS Inc., Chicago), SAS Version 9.2 (SAS Institute Inc., Cary, NC), and Winsteps Version 3.75.0 (Linacre, 2012). Phase 1 analysis used descriptive statistics, principal component analysis, and correlation analysis with the intention of identifying problematic items and culling or rewriting them. We were also intent on ensuring that items did not strongly correlate with either gender or substance so that the end product would be gender and substance neutral.
Principal component analysis revealed strong unidimensionality of the instrument, as illustrated in Figure 1. We identified 32 “factors” or variables (item clusters) with eigenvalues above 1.0 that represent classification axes. The coordinates of factors, sometimes referred to as the factor loadings, represent the strength of the relationship between each item and the various factors (Field, 2005). Principal component analysis can produce as many factors as there are items, but not all are necessarily of substantive importance (Field, 2005). The scree plot of eigenvalues (Figure 1) has a typically sharp decline in the curve and then levels off, giving the classical elbow-shaped configuration. Conventional reading of principal component analysis involves accepting all factors with eigenvalues of 1.0 or higher (Kaiser, 1960); however, there is also support for cutting off the number of factors on the basis of the point the scree plot levels off and researcher evaluation (Cattell, 1966).
Figure 1.
Scree plots for Phase 1 (top) and Phase 2 (bottom) testing.
The classic elbow-shaped scree plot is indicative of one unifying construct, with major factors indicated on the downward slope.
Figure 1.
Scree plots for Phase 1 (top) and Phase 2 (bottom) testing.
The classic elbow-shaped scree plot is indicative of one unifying construct, with major factors indicated on the downward slope.
×
Correlational analysis of items revealed significant relationships between some of the items and the demographic factors of gender and type of substance (drug or alcohol). These items were examined and either removed from the instrument or revised to remove any obvious link to those factors. After revision of the instrument, only two items showed correlations with these factors above .20; these relationships were weak, so the items were retained.
Examination of the demographic items revealed that length of time abusing substances had serious ceiling effects; we expanded the options to cover a longer period of years. Some respondents had trouble answering the marital status questions, not sure which category they fit, so we expanded that list to include living with a significant other. Finally, we found it necessary to expand the preferred substance item by changing the drug category to “prescription drugs” and “nonprescription drugs”; some participants abused prescription drugs they had been prescribed or took pills from supplies prescribed to relatives or others. The resulting revised instrument used in Phase 2 consisted of 70 items and an expanded or refined demographics page.
Phase 2 Data Collection and Analysis
In Phase 2, we used the revised instrument consisting of 70 items and the demographic variables with expanded or refined items regarding length of addiction, marital status, and substance preference. Data collection occurred over a span of 5 mo with participants who were at the end of their detoxification program or near the beginning of residential treatment. The same two treatment centers in southwest Florida participated. The Phase 2 sample consisted of 276 people, of whom 268 completed questionnaires. Participants were between ages 19 and 61 yr; 60.1% were male; and 81.6% were White, 5.6% Hispanic, 4.9% African-American, 3.0% Native American, and 4.9% other or mixed. On the expanded scale for length of substance dependence, 53.4% reported 10 or more years of dependency, 13.1% reported 6–9 yr, 21.3% reported 3–5 yr, and 12.2% reported no more than 2 yr.
Data analysis was done using IBM SPSS Statistics (Version 20; IBM Corp., Armonk, NY), SAS Version 9.2, and Winsteps Version 3.75.0. Descriptive statistics were used to examine the properties of the instrument at the individual item, subscore, and total score levels; α coefficients and Pearson correlations were also examined. Psychometric properties were further examined using exploratory and confirmatory factor analysis; principal axis factor analysis was run using a varimax rotation for the total revised instrument to examine the dimensionality and to extract meaningful rotated factor structures. The factor analytical results were also used to examine the fit of each item both to the entire instrument and to each factor. Item response theory (IRT) calibrations using the Andrich (1978)  extension of the Rasch family of models (Wright & Stone, 1979) were also used because they are most appropriate for evaluation of rating scale data. This analysis allowed us to begin looking at model fit and scaling of the instrument.
Results
Item, subfactor, and total means and standard deviations from Phase 2 data are shown in Table 1; correlations between all the factors can be seen in Table 2. Factor analysis was run to examine the results for both exploratory and confirmatory purposes. Exploratory factor analysis was first examined to allow us to see the loadings for all items with no restrictions placed on the model. In the next factor analysis, we placed a priori constraint on the model using occupational therapy and White’s conceptual models. We examined both of these analyses to determine a best fitting model using data and expert opinion. Although we identified 13 factors with eigenvalues above 1.0, the scree plot (see Figure 1) supported limiting the number of factors to 8 (see Table 1).
Table 1.
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire×
FactorDescriptive StatisticsInitial Eigenvalues
NMinMaxMeanSDTotal% of VarianceCumulative %
1. Occupational Disruption26816441.3614.33816.98324.26224.262
2. Habits and Routines26834834.859.5104.0245.74930.011
3. Social Environment26802415.535.2103.6655.23635.246
4. Family Disapprobation26801612.143.6783.2464.63739.884
5. Residual Strengths2670248.583.7762.3693.38443.267
6. Self-Medicating Behaviors2680129.402.9881.9552.79246.060
7. Physical Environment2680123.442.6751.8672.66848.728
8. Readiness for Change268082.121.7331.7902.55751.285
 Total26846228165.7536.28
Table Footer NoteNote. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.
Note. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.×
Table 1.
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire×
FactorDescriptive StatisticsInitial Eigenvalues
NMinMaxMeanSDTotal% of VarianceCumulative %
1. Occupational Disruption26816441.3614.33816.98324.26224.262
2. Habits and Routines26834834.859.5104.0245.74930.011
3. Social Environment26802415.535.2103.6655.23635.246
4. Family Disapprobation26801612.143.6783.2464.63739.884
5. Residual Strengths2670248.583.7762.3693.38443.267
6. Self-Medicating Behaviors2680129.402.9881.9552.79246.060
7. Physical Environment2680123.442.6751.8672.66848.728
8. Readiness for Change268082.121.7331.7902.55751.285
 Total26846228165.7536.28
Table Footer NoteNote. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.
Note. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.×
×
Table 2.
Correlations Between Factors
Correlations Between Factors×
FactorMeasure2345678
1a. Occupational DisruptionPearson correlation.651**.568**.600**.032.434**.380**−.289**
p.000.000.000.605.000.000.000
N268268268267268268268
2a. Habits and RoutinesPearson correlation.562**.581**−.047.713**.325**−.234**
p.000.000.440.000.000.000
N268268267268268268
3a. Social EnvironmentPearson correlation.448**−.040.408**.455**−.173**
p.000.513.000.000.005
N268267268268268
4a. Family DisapprobationPearson correlation−.109.476**.264**−.350**
p.076.000.000.000
N267268268268
5a. Residual StrengthsPearson correlation−.064−.021.197**
p.299.729.001
N267267267
6a. Self-Medicating BehaviorsPearson correlation.236**−.142*
p.000.020
N268268
7a. Readiness for ChangePearson correlation−.066
p.285
N268
Table Footer Note*Correlation is significant at the .05 level (2-tailed).
Correlation is significant at the .05 level (2-tailed).×
Table Footer Note**Correlation is significant at the .01 level (2-tailed).
Correlation is significant at the .01 level (2-tailed).×
Table 2.
Correlations Between Factors
Correlations Between Factors×
FactorMeasure2345678
1a. Occupational DisruptionPearson correlation.651**.568**.600**.032.434**.380**−.289**
p.000.000.000.605.000.000.000
N268268268267268268268
2a. Habits and RoutinesPearson correlation.562**.581**−.047.713**.325**−.234**
p.000.000.440.000.000.000
N268268267268268268
3a. Social EnvironmentPearson correlation.448**−.040.408**.455**−.173**
p.000.513.000.000.005
N268267268268268
4a. Family DisapprobationPearson correlation−.109.476**.264**−.350**
p.076.000.000.000
N267268268268
5a. Residual StrengthsPearson correlation−.064−.021.197**
p.299.729.001
N267267267
6a. Self-Medicating BehaviorsPearson correlation.236**−.142*
p.000.020
N268268
7a. Readiness for ChangePearson correlation−.066
p.285
N268
Table Footer Note*Correlation is significant at the .05 level (2-tailed).
Correlation is significant at the .05 level (2-tailed).×
Table Footer Note**Correlation is significant at the .01 level (2-tailed).
Correlation is significant at the .01 level (2-tailed).×
×
After carefully examining the items clustered in each factor, we identified eight factors: Factor 1, Occupational Disruption; Factor 2, Habits and Routines; Factor 3, Social Environment (which reflects cultural enmeshment as described by White); Factor 4, Family Disapprobation (which is related to social context and cultural enmeshment); Factor 5, Residual Strengths; Factor 6, Self-Medicating Behaviors; Factor 7, Physical Environment; and Factor 8, Readiness for Change. As shown in Table 2, high positive correlations between all factors were found, with the exception of Factor 5, Residual Strengths, and Factor 8, Readiness for Change, in which the correlations were either low or negative. Factors 5 and 8 contained the items that were reverse scored to maintain directionality of the instrument; negative correlations were therefore expected.
Reliability Evidence
The total instrument α coefficient was .93. The subscale αs and number of items in the subscales were as follows: Factor 1, Occupational Disruption, α = .92 (15 items); Factor 2, Habits and Routines, α = .87 (11 items); Factor 3, Social Environment, α = .82 (6 items); Factor 4, Family Disapprobation, α = .83 (4 items); Factor 5, Residual Strengths, α = .53 (6 items); Factor 6, Self-Medicating Behaviors, α = .80 (3 items); Factor 7, Physical Environment, α = .71 (3 items); and Factor 8, Readiness for Change, α = .70 (2 items). The most notable reliability findings were the high coefficients for those scales that contain only 2, 3, or 4 items and the somewhat lower coefficient for Factor 5, Residual Strengths. Items not included in these factors may have correlated with one or more factors, but the correlations were not considered strong enough to include in a subscale.
Total instrument reliability from the IRT calibration analyses for person reliability was .92 and for item reliability was .99. Person reliability addresses the issue of whether the instrument consistently detects sufficient levels of difference among people within the sample; high person reliability indicates a high probability that differences in performance on the instrument reflect actual differences in the respondents’ experiences. Item reliability refers to whether the instrument consistently reflects sufficient variance in levels of difficulty (or variation in responses). High item reliability indicates a high probability that high and low scores on the instrument reflect actual performance (experience) levels of participants; low item reliability also may result from a small sample (Wright & Masters, 1982). Our findings show both high person reliability and high item reliability and provide additional evidence that the total score reliability is high.
Validity Evidence
The validity evidence examined included confirmatory factor analysis of the structure of the subscales and expert examination of the items, substructures, and total structure of the instrument. The scree plots from both phases of factor analysis (Figure 1) illustrate the unidimensionality of the instrument and further confirm our contention that the construct is cohesive. Correlations of the subscores were also examined to posit the relationships or lack thereof among the subscores (Table 2). The correlational results produced moderate to strong relationships ranging from −.35 to .71; 11 of the correlations were >.40. These data provide additional construct validity evidence that each factor subscore provides some common and some unique variance. Examination of the factors that emerged informed our decisions regarding domain score names and the naming of the unifying construct, Occupational Dysfunction.
To ascertain the relationship between gender, ethnicity, and type of drug used, and thus to provide validity evidence that the structures of the subscores are consistent for both genders, all ethnic groups, and all types of drug used, we conducted analyses of variance and t tests using the demographic variables (gender, age, race and ethnicity, marital status, education level, preferred substance, and length of addiction) as the independent variables and the total scores as the dependent variable. Nonsignificant results of these comparisons would provide validity evidence that the instrument obtains consistent results for some of the more typical subpopulations and thus that the construct remains invariant across groups. The group comparisons were not significantly different, with the exception of a gender difference on the total instrument score; the female group mean (172.68, standard deviation [SD] = 35.576, n = 107) was significantly higher than the male group mean (161.14, SD = 36.148, n = 161), t (266) = 2.58, p < .01.
Item Response Theory Calibration and Analysis
To further examine the fit and dimensionality of the items and therefore the construct validity, we calibrated the item responses with Winsteps Version 3.75.0 using the Andrich (1978)  rating scale IRT model. This model is a polytomous extension of the Rasch one-parameter model (Wright & Stone, 1979), and it is appropriate when rating scale item types are being calibrated. We examined the item and person statistics with a focus on the mean square infit and outfit statistics for each item. Typically, for good item fit, fit values should be between 0.60 and 1.40. Linacre (2002), however, provided guidelines for interpretation of mean square values that vary slightly from this standard and are less stringent. He suggested the following: >2.0 distorts or degrades the measurement system, 1.5–2.0 is unproductive for construction of the measurement system but not degrading, 0.5–1.5 is productive for the measurement system, and <0.5 is less productive for the measurement system, is not degrading, but may produce misleadingly high reliability and separation coefficients.
On examination of the data, we found that two of the mean square values were degrading (>2.0) and that 8 items had infit or outfit mean square values between 1.5 and 2.0. The items that exceeded these thresholds were not at the extremes (i.e., top or bottom) of the scale. Extreme values usually have high or low fit statistics by definition. We used the overall summary results from the calibration (see Table 3) to assess dimensionality and to provide several more pieces of information for use in confirming the construct and dimensionality of the instrument. Detailed results from the calibration (not included here because of space constraints) will inform future decisions regarding further revision of the instrument.
Table 3.
Summary of Results From Item Response Theory Calibration
Summary of Results From Item Response Theory Calibration×
ValueTotal ScoreCountMeasureModel ErrorInfitOutfit
MnsqZstdMnsqZstd
Mean165.768.90.290.111.01−0.21.06−0.1
SD36.23.60.440.010.432.50.662.4
Maximum228.070.01.240.163.126.98.009.9
Minimum46.028.0−1.220.100.28−6.60.29−5.8
RMSE
True SD
Separation
Person Reliability
Real RMSE.12.423.35.92
Model RMSE.12.423.65.93
Table Footer NoteNote. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).
Note. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).×
Table 3.
Summary of Results From Item Response Theory Calibration
Summary of Results From Item Response Theory Calibration×
ValueTotal ScoreCountMeasureModel ErrorInfitOutfit
MnsqZstdMnsqZstd
Mean165.768.90.290.111.01−0.21.06−0.1
SD36.23.60.440.010.432.50.662.4
Maximum228.070.01.240.163.126.98.009.9
Minimum46.028.0−1.220.100.28−6.60.29−5.8
RMSE
True SD
Separation
Person Reliability
Real RMSE.12.423.35.92
Model RMSE.12.423.65.93
Table Footer NoteNote. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).
Note. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).×
×
One of the advantages of the IRT calibration statistics is that they are person invariant and thus are not influenced by the fact that the sample was one of convenience. Given the property of person invariance, the IRT results revealed information that was different from all of the classical analyses (sample dependent) we reviewed during the item examination meetings.
Discussion
Our purposes in developing the LHQ were to create a relatively brief but sound measure to gather information about patterns of clients’ lives in addiction that would contribute to individualized treatment interventions. We were also interested in whether factor analysis supported White’s conceptualizations regarding the categories of experience and performance levels, patterns, and environmental influence that occupational therapy practitioners would expect.
What we did not expect was the result showing so many factors with eigenvalues above 1.0. The factors not included in the final eight comprised 1 or 2 items or a few items with fairly weak factor loadings. This fact, along with the scree plot evidence, was the basis for limiting the identified factors to eight.
The results of this study support the validity and reliability of the LHQ for measuring the degree of occupational dysfunction of people living with substance addiction. Data analysis using a number of approaches showed a psychometrically sound instrument. Factor analysis supported a strong one-factor solution (Occupational Dysfunction) and also supported White’s fundamental construct of cultural enmeshment, which occupational therapy practitioners typically characterize as part of the social environment.
Factor 1, Occupational Disruption, reflects performance in a broad spectrum of areas of occupation (activities of daily living, instrumental activities of daily living, rest and sleep, education, work, and leisure), and some items strongly suggest erosion of what Kielhofner (2002)  described as the volitional subsystem (e.g., Item 35, “I felt unmotivated to do certain activities,” and Item 44, “I didn’t have much interest in everyday activities”). Performance patterns are represented by Factor 2, Habits and Routines. Performance mediated by the social environment (White’s cultural enmeshment) is captured in Factor 3, Social Environment, and Factor 4, Family Disapprobation, is also related to social environment and cultural enmeshment. The reliability estimate for each of these factors is strong.
Limitations and Future Research
The sample size for both phases of this project was adequate, but a large segment of both samples were people with chronic substance addictions within a restricted geographic area. Significantly more men than women were represented, and ethnicity was overwhelmingly White and non-Hispanic. Further testing of the instrument on a more representative sample in terms of gender, race and ethnicity, and length of dependence on substances is warranted. Evaluation of test–retest reliability of the instrument, also essential, is currently in the planning stages.
Further study of this instrument should include a closer look at the items that lie outside the ideal mean square value range to determine whether they should be revised or deleted from the instrument. Likewise, the development of scaled scores in the various domains and client profile scores would add to the utility of the instrument. A revised, shortened version of the LHQ is now in development to improve its potential for use in the current health care environment. An adolescent version is also in the planning stages.
Whether the instrument will prove useful in the clinic for evaluation or for research has yet to be determined. Preliminary observations suggest that using the instrument to measure outcomes of treatment while the person is still living in a treatment community (halfway house or supported housing) is of transient value because floor effects are pronounced—people are insulated from the environments that have contributed to prior patterns of substance use, so scores based on recent experience plummet. That said, the instrument might provide an outcome measure that is the new baseline measure for people returning to their prior environment or any environment without the structure of a treatment community. The instrument may also be useful for following clients who remain in their home and community during treatment or who return to their original community after treatment. More study of the instrument and its applications is necessary to determine the suitability of its use as an outcome measure.
Implications for Occupational Therapy Practice
  • The LHQ is a promising new assessment of occupational dysfunction for people entering treatment for substance abuse.

  • The LHQ contains factors that reflect the concerns of occupational therapy practitioners with clients’ performance in context.

  • The instrument may be an efficient way to gather information regarding clients’ history and performance patterns, but its use should be considered experimental at present, and practitioners should not rely solely on the LHQ for client evaluation.

Conclusion
Occupational disturbances are a central feature of living with substance addictions. During data collection for this study, many participants stopped to talk with us, commenting that completing the measure made them realize the extent to which their lives had been transformed or “highjacked” by substance abuse. Occupational therapy practitioners recognize this as a central truth, and the results of our statistical analyses provide strong evidence to inform interventions that address problems of performance in context.
Essential to the treatment of substance addictions is the realization that habits and routines, beyond the physiological drug or alcohol dependence, are deeply ingrained in the subconscious mind. These maladaptive habits develop to accommodate the addiction or as a part of the addiction. On entering treatment, clients experience gaps in their routines that were formerly filled by substance use, and the absence of a new routine can trigger relapse. Extensive repetition of new replacement habits and patterns of task performance, along with contextual change, is required to make the transition to recovery. Much of current substance abuse treatment focuses on changing the person and his or her thinking and commitment to change, which are important. The strength of occupational therapy is its focus on the performance and contextual elements that have great potential to promote long-term recovery.
Acknowledgments
The authors thank the staff of the Transitional Living Center and the Detoxification Unit of Southwest Florida Addiction Services (now merged with Lee Mental Health in a new organization, SalusCare, Inc.) for their contributions to the project and to the management of the organization for the care and attention given to protection of participants’ privacy. Special thanks go to Alex Garcia Barbon, assistant vice president for acute care services, for his assistance in providing access and time for our recruitment of participants and gathering of data.
References
American Occupational Therapy Association. (2014). Occupational therapy practice framework: Domain and process (3rd ed.). American Journal of Occupational Therapy, 68(Suppl. 1), S1−S48. http://dx.doi.org/10.5014.ajot.2014.682006 [Article]
American Occupational Therapy Association. (2014). Occupational therapy practice framework: Domain and process (3rd ed.). American Journal of Occupational Therapy, 68(Suppl. 1), S1−S48. http://dx.doi.org/10.5014.ajot.2014.682006 [Article] ×
Andrich, D. A. (1978). A rating formulation for ordered response categories. Psychometrika, 43, 561–573. http://dx.doi.org/10.1007/BF02293814 [Article]
Andrich, D. A. (1978). A rating formulation for ordered response categories. Psychometrika, 43, 561–573. http://dx.doi.org/10.1007/BF02293814 [Article] ×
Brown, C. (2014). Ecological models in occupational therapy. In B. A. B. Schell, G. Gillen, & M. Scaffa (Eds.), Willard and Spackman’s occupational therapy (12th ed., pp. 494−504). Baltimore: Lippincott Williams & Wilkins.
Brown, C. (2014). Ecological models in occupational therapy. In B. A. B. Schell, G. Gillen, & M. Scaffa (Eds.), Willard and Spackman’s occupational therapy (12th ed., pp. 494−504). Baltimore: Lippincott Williams & Wilkins.×
Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245–276. http://dx.doi.org/10.1207/s15327906mbr0102_10 [Article]
Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245–276. http://dx.doi.org/10.1207/s15327906mbr0102_10 [Article] ×
Christiansen, C., & Baum, C. (Eds.). (1997). Occupational therapy: Enabling function and well-being (2nd ed.). Thorofare, NJ: Slack.
Christiansen, C., & Baum, C. (Eds.). (1997). Occupational therapy: Enabling function and well-being (2nd ed.). Thorofare, NJ: Slack.×
Dunn, W., Brown, C., & McGuigan, A. (1994). The ecology of human performance: A framework for considering the effect of context. American Journal of Occupational Therapy, 48, 595–607. http://dx.doi.org/10.5014/ajot.48.7.595 [Article] [PubMed]
Dunn, W., Brown, C., & McGuigan, A. (1994). The ecology of human performance: A framework for considering the effect of context. American Journal of Occupational Therapy, 48, 595–607. http://dx.doi.org/10.5014/ajot.48.7.595 [Article] [PubMed]×
Field, A. (2005). Discovering statistics using SPSS (2nd ed.). London: Sage.
Field, A. (2005). Discovering statistics using SPSS (2nd ed.). London: Sage.×
Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20, 141–151. http://dx.doi.org/10.1177/001316446002000116 [Article]
Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20, 141–151. http://dx.doi.org/10.1177/001316446002000116 [Article] ×
Kielhofner, G. (2002). A model of human occupation: Theory and application. Baltimore: Lippincott Williams & Wilkins.
Kielhofner, G. (2002). A model of human occupation: Theory and application. Baltimore: Lippincott Williams & Wilkins.×
Kielhofner, G., Mallinson, T., Crawford, C., Nowak, M., Rigby, M., Henry, A., & Walens, D. (1998). A user’s manual for the Occupational Performance History Interview (OPHI−II), Version 2.1. Chicago: University of Illinois at Chicago.
Kielhofner, G., Mallinson, T., Crawford, C., Nowak, M., Rigby, M., Henry, A., & Walens, D. (1998). A user’s manual for the Occupational Performance History Interview (OPHI−II), Version 2.1. Chicago: University of Illinois at Chicago.×
Law, M., Baptiste, S., Carswell, A., McColl, M. A., Polatajko, H., & Pollock, N. (2005). Canadian Occupational Performance Measure (4th ed.). Ottawa: CAOT Publications ACE.
Law, M., Baptiste, S., Carswell, A., McColl, M. A., Polatajko, H., & Pollock, N. (2005). Canadian Occupational Performance Measure (4th ed.). Ottawa: CAOT Publications ACE.×
Law, M., Cooper, B., Strong, S., Steward, D., Rigby, P., & Letts, L. (1996). The Person−Environment−Occupation model: A transactive approach to occupational performance. Canadian Journal of Occupational Therapy, 63, 9–23. http://dx.doi.org/10.1177/000841749606300103 [Article]
Law, M., Cooper, B., Strong, S., Steward, D., Rigby, P., & Letts, L. (1996). The Person−Environment−Occupation model: A transactive approach to occupational performance. Canadian Journal of Occupational Therapy, 63, 9–23. http://dx.doi.org/10.1177/000841749606300103 [Article] ×
Lemke, E., & Wiersma, W. (1976). Principles of psychological measurement. Chicago: Rand McNally.
Lemke, E., & Wiersma, W. (1976). Principles of psychological measurement. Chicago: Rand McNally.×
Linacre, M. (2002). What do infit and outfit, mean-square and standardized mean? Rasch Measurement Transactions, 16(2), 878. Retrieved from http://www.rasch.org/rmt/rmt162f.htm
Linacre, M. (2002). What do infit and outfit, mean-square and standardized mean? Rasch Measurement Transactions, 16(2), 878. Retrieved from http://www.rasch.org/rmt/rmt162f.htm×
Linacre, M. (2012). Winsteps® Rasch measurement computer program. Beaverton, OR: Winsteps.
Linacre, M. (2012). Winsteps® Rasch measurement computer program. Beaverton, OR: Winsteps.×
McLellan, A. T., Luborsky, L., Woody, G. E., & O’Brien, C. P. (1980). An improved diagnostic evaluation instrument for substance abuse patients: The Addiction Severity Index. Journal of Nervous and Mental Disease, 168, 26–33. http://dx.doi.org/10.1097/00005053-198001000-00006 [Article] [PubMed]
McLellan, A. T., Luborsky, L., Woody, G. E., & O’Brien, C. P. (1980). An improved diagnostic evaluation instrument for substance abuse patients: The Addiction Severity Index. Journal of Nervous and Mental Disease, 168, 26–33. http://dx.doi.org/10.1097/00005053-198001000-00006 [Article] [PubMed]×
Neal, D. T., Wood, W., & Quinn, J. M. (2006). Habits—A repeat performance. Current Directions in Psychological Science, 15, 198–202. http://dx.doi.org/10.1111/j.1467-8721.2006.00435.x [Article]
Neal, D. T., Wood, W., & Quinn, J. M. (2006). Habits—A repeat performance. Current Directions in Psychological Science, 15, 198–202. http://dx.doi.org/10.1111/j.1467-8721.2006.00435.x [Article] ×
Schkade, J. K., & Schultz, S. (1992). Occupational adaptation: Toward a holistic approach for contemporary practice, Part 1. American Journal of Occupational Therapy, 46, 829–837. http://dx.doi.org/10.5014/ajot.46.9.829 [Article] [PubMed]
Schkade, J. K., & Schultz, S. (1992). Occupational adaptation: Toward a holistic approach for contemporary practice, Part 1. American Journal of Occupational Therapy, 46, 829–837. http://dx.doi.org/10.5014/ajot.46.9.829 [Article] [PubMed]×
Streiner, D. L. (2003). Starting at the beginning: An introduction to coefficient alpha and internal consistency. Journal of Personality Assessment, 80, 99–103. http://dx.doi.org/ 10.1207/S15327752JPA8001_18 [Article] [PubMed]
Streiner, D. L. (2003). Starting at the beginning: An introduction to coefficient alpha and internal consistency. Journal of Personality Assessment, 80, 99–103. http://dx.doi.org/ 10.1207/S15327752JPA8001_18 [Article] [PubMed]×
White, W. L. (1996). Pathways from the culture of addiction to the culture of recovery: A travel guide for addiction professionals (2nd ed.). Center City, MN: Hazelden Press.
White, W. L. (1996). Pathways from the culture of addiction to the culture of recovery: A travel guide for addiction professionals (2nd ed.). Center City, MN: Hazelden Press.×
White, W. L. (2009). The mobilization of community resources to support long-term addiction recovery. Journal of Substance Abuse Treatment, 36, 146–158. http://dx.doi.org/10.1016/j.jsat.2008.10.006 [Article] [PubMed]
White, W. L. (2009). The mobilization of community resources to support long-term addiction recovery. Journal of Substance Abuse Treatment, 36, 146–158. http://dx.doi.org/10.1016/j.jsat.2008.10.006 [Article] [PubMed]×
Wood, W., Tam, L., & Witt, M. G. (2005). Changing circumstances, disrupting habits. Journal of Personality and Social Psychology, 88, 918–933. http://dx.doi.org/10.1037/0022-3514.88.6.918 [Article] [PubMed]
Wood, W., Tam, L., & Witt, M. G. (2005). Changing circumstances, disrupting habits. Journal of Personality and Social Psychology, 88, 918–933. http://dx.doi.org/10.1037/0022-3514.88.6.918 [Article] [PubMed]×
Wright, B., & Masters, G. (1982). Rating scale analysis. Chicago: Mesa Press.
Wright, B., & Masters, G. (1982). Rating scale analysis. Chicago: Mesa Press.×
Wright, B. D., & Stone, M. H. (1979). Best test design. Chicago: University of Chicago Social Research. Available at http://www.rasch.org/btd.htm
Wright, B. D., & Stone, M. H. (1979). Best test design. Chicago: University of Chicago Social Research. Available at http://www.rasch.org/btd.htm×
Figure 1.
Scree plots for Phase 1 (top) and Phase 2 (bottom) testing.
The classic elbow-shaped scree plot is indicative of one unifying construct, with major factors indicated on the downward slope.
Figure 1.
Scree plots for Phase 1 (top) and Phase 2 (bottom) testing.
The classic elbow-shaped scree plot is indicative of one unifying construct, with major factors indicated on the downward slope.
×
Table 1.
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire×
FactorDescriptive StatisticsInitial Eigenvalues
NMinMaxMeanSDTotal% of VarianceCumulative %
1. Occupational Disruption26816441.3614.33816.98324.26224.262
2. Habits and Routines26834834.859.5104.0245.74930.011
3. Social Environment26802415.535.2103.6655.23635.246
4. Family Disapprobation26801612.143.6783.2464.63739.884
5. Residual Strengths2670248.583.7762.3693.38443.267
6. Self-Medicating Behaviors2680129.402.9881.9552.79246.060
7. Physical Environment2680123.442.6751.8672.66848.728
8. Readiness for Change268082.121.7331.7902.55751.285
 Total26846228165.7536.28
Table Footer NoteNote. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.
Note. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.×
Table 1.
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire
Descriptive Statistics and Eigenvalues of the Eight Factors of the Lifestyle History Questionnaire×
FactorDescriptive StatisticsInitial Eigenvalues
NMinMaxMeanSDTotal% of VarianceCumulative %
1. Occupational Disruption26816441.3614.33816.98324.26224.262
2. Habits and Routines26834834.859.5104.0245.74930.011
3. Social Environment26802415.535.2103.6655.23635.246
4. Family Disapprobation26801612.143.6783.2464.63739.884
5. Residual Strengths2670248.583.7762.3693.38443.267
6. Self-Medicating Behaviors2680129.402.9881.9552.79246.060
7. Physical Environment2680123.442.6751.8672.66848.728
8. Readiness for Change268082.121.7331.7902.55751.285
 Total26846228165.7536.28
Table Footer NoteNote. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.
Note. — = not applicable; Max = maximum; Min = minimum; SD = standard deviation.×
×
Table 2.
Correlations Between Factors
Correlations Between Factors×
FactorMeasure2345678
1a. Occupational DisruptionPearson correlation.651**.568**.600**.032.434**.380**−.289**
p.000.000.000.605.000.000.000
N268268268267268268268
2a. Habits and RoutinesPearson correlation.562**.581**−.047.713**.325**−.234**
p.000.000.440.000.000.000
N268268267268268268
3a. Social EnvironmentPearson correlation.448**−.040.408**.455**−.173**
p.000.513.000.000.005
N268267268268268
4a. Family DisapprobationPearson correlation−.109.476**.264**−.350**
p.076.000.000.000
N267268268268
5a. Residual StrengthsPearson correlation−.064−.021.197**
p.299.729.001
N267267267
6a. Self-Medicating BehaviorsPearson correlation.236**−.142*
p.000.020
N268268
7a. Readiness for ChangePearson correlation−.066
p.285
N268
Table Footer Note*Correlation is significant at the .05 level (2-tailed).
Correlation is significant at the .05 level (2-tailed).×
Table Footer Note**Correlation is significant at the .01 level (2-tailed).
Correlation is significant at the .01 level (2-tailed).×
Table 2.
Correlations Between Factors
Correlations Between Factors×
FactorMeasure2345678
1a. Occupational DisruptionPearson correlation.651**.568**.600**.032.434**.380**−.289**
p.000.000.000.605.000.000.000
N268268268267268268268
2a. Habits and RoutinesPearson correlation.562**.581**−.047.713**.325**−.234**
p.000.000.440.000.000.000
N268268267268268268
3a. Social EnvironmentPearson correlation.448**−.040.408**.455**−.173**
p.000.513.000.000.005
N268267268268268
4a. Family DisapprobationPearson correlation−.109.476**.264**−.350**
p.076.000.000.000
N267268268268
5a. Residual StrengthsPearson correlation−.064−.021.197**
p.299.729.001
N267267267
6a. Self-Medicating BehaviorsPearson correlation.236**−.142*
p.000.020
N268268
7a. Readiness for ChangePearson correlation−.066
p.285
N268
Table Footer Note*Correlation is significant at the .05 level (2-tailed).
Correlation is significant at the .05 level (2-tailed).×
Table Footer Note**Correlation is significant at the .01 level (2-tailed).
Correlation is significant at the .01 level (2-tailed).×
×
Table 3.
Summary of Results From Item Response Theory Calibration
Summary of Results From Item Response Theory Calibration×
ValueTotal ScoreCountMeasureModel ErrorInfitOutfit
MnsqZstdMnsqZstd
Mean165.768.90.290.111.01−0.21.06−0.1
SD36.23.60.440.010.432.50.662.4
Maximum228.070.01.240.163.126.98.009.9
Minimum46.028.0−1.220.100.28−6.60.29−5.8
RMSE
True SD
Separation
Person Reliability
Real RMSE.12.423.35.92
Model RMSE.12.423.65.93
Table Footer NoteNote. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).
Note. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).×
Table 3.
Summary of Results From Item Response Theory Calibration
Summary of Results From Item Response Theory Calibration×
ValueTotal ScoreCountMeasureModel ErrorInfitOutfit
MnsqZstdMnsqZstd
Mean165.768.90.290.111.01−0.21.06−0.1
SD36.23.60.440.010.432.50.662.4
Maximum228.070.01.240.163.126.98.009.9
Minimum46.028.0−1.220.100.28−6.60.29−5.8
RMSE
True SD
Separation
Person Reliability
Real RMSE.12.423.35.92
Model RMSE.12.423.65.93
Table Footer NoteNote. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).
Note. Mnsq = mean square; RMSE = root mean square error; SD = standard deviation; Zstd = Z-standardized. Valid responses: 98.4% (approximate). Standard error of person mean = .03. Person raw score–to–measure correlation = .97 (approximate because of missing data). Cronbach’s α (Kuder–Richardson Formula 20) person raw score reliability = .94 (approximate because of missing data).×
×