Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Sep 1.
Published in final edited form as: Behav Ther. 2024 Feb 29;55(5):990–1003. doi: 10.1016/j.beth.2024.02.003

Psychometric Evaluation of the Insomnia Severity Index in U.S. College Students

Sarah E Emert 1, Jessica R Dietch 2, Adam D Bramoweth 3, Kimberly Kelly 4, Daniel J Taylor 1
PMCID: PMC11341948  NIHMSID: NIHMS1971885  PMID: 39174275

Abstract

Psychometric properties of the Insomnia Severity Index (ISI) were analyzed in U.S. college samples. ISI items and total score with sleep and psychosocial questionnaires were examined in Experiment I. ISI diagnostic accuracy in a clinical sample with and without insomnia was assessed in Experiment II. ISI test-retest validity, confirmatory factor analysis (CFA), and Item Response Theory via Graded Response Model (GRM) were assessed in Experiment III. Results indicated analogous ISI and sleep diary items showed moderate correlations (r1 = .40; r2 = .45). The ISI total had weak to strong correlations with other indicators of sleep-related disturbance (rs = .25 - .62). The ISI had weak to moderate correlations with psychosocial measures commonly associated with insomnia (rs = .10 - .57). The diagnostic accuracy of the ISI was very high (area under the curve = .999). Sensitivity and specificity were maximized at a cutoff score ≥ 8. The ISI demonstrated good test-retest reliability (ICC = .87). CFA revealed a three-factor model for two study samples and GRM indicated better ability of the ISI to assess moderate (Sample III) and moderate to high (Sample I) levels of insomnia severity. The ISI demonstrated good psychometric properties and appears generally valid for screening insomnia disorder and assessing insomnia severity in college students. Overlap with psychological symptoms suggests caution while interpreting these constructs independently.

Keywords: Insomnia Severity Index, College Students, Psychometrics, Validation, Reliability

Introduction

The transition experienced by students during college years may precipitate insomnia complaints (e.g., 16-19.3%; Albasheer et al., 2020; Taylor et al., 2013) and subclinical symptomatology (e.g., 60%; Schlarb et al., 2017) which are common among this population. Determining the validity of measures to accurately screen and identify insomnia disorder during this crucial social and developmental timeframe is imperative, because the consequences of insomnia in this population include increased risk for worse quality of life and mental health (e.g., suicide, depression, anxiety, stress), increased hypnotic and stimulant use, and cognitive difficulties (Roane & Taylor, 2008; Taylor et al., 2013; Wilkerson et al., 2012; Yang et al., 2003).

The Insomnia Severity Index (ISI) is a brief self-report questionnaire that is the most commonly used self-report measure for the clinical evaluation of insomnia (Bastien et al., 2001) and has been validated in numerous nationalities, languages, and medical populations (Chahoud et al., 2017; Dragioti et al., 2015; Gagnon et al., 2013; Kaufmann et al., 2019; Lin et al., 2020; Morin; Moscou-Jackson et al., 2016). Although previously validated in adolescents and young adults (Chahoud et al., 2017; Chung et al., 2011; Gerber et al., 2016) and international college populations (Gerber et al., 2016; Veqar & Hussain, 2017; Wong et al., 2017) no studies have examined the psychometric properties of the ISI in a U.S. college sample.

The goal of these analyses was to evaluate the psychometric properties of the ISI within three samples of students at a public university in the south-central United States. These studies are essential in demonstrating the ISI measures the same construct (i.e., insomnia) in the same way in college students (e.g., internal and external validity, and reliability), which ensures generalizability and ecological validity of other findings in this important population (i.e., 62% of the U.S. population attends college after high school or GED (U.S. Department of Commerce, 2022). With these goals in mind, we assessed: the overlap of the ISI items and total score with other sleep and psychosocial measures; the diagnostic accuracy of the ISI for insomnia disorder; the reliability of the ISI across time; factor structure and among item-level comparisons.

Experiment I

The aim of Experiment I was to examine ISI items and total score to well-validated measures of sleep and psychosocial functioning.

Methods

Sample I

This secondary analysis utilized previously collected data (Fall 2006-Spring 2007). A total of 1,040 undergraduate students volunteered to complete study questionnaires in exchange for psychology course credit. The final sample size was 1,040 students with an average age of 20.38 years (SD = 3.93) and 71.4% were females (4.2% did not provide sex information). Sample characteristics are presented in Table 1. The sample was predominantly non-Hispanic (89.6%) and white (74.9%) and the racial/ethnic makeup was similar to the university. This study was approved by the university institutional review board and informed consent was obtained from all participants.

Table 1.

Characteristics for Samples I, II, and III

Sample I (N=1,040) Sample II (N=153) Sample III (N=395)
n (%) or mean ± SD n (%) or mean ± SD n (%) or mean ± SD
Age 20.38 ± 3.93 20.24 ± 2.57 20.79 ± 3.44
Sex
Male 253 (24.3) 61 (39.9) 106 (25.9)
Female 743 (71.4) 88 (57.5) 286 (69.9)
Transgender male -- -- 1 (0.3)
Genderqueer -- -- 1 (0.3)
Race
American Indian/Alaska Native -- -- 5 (1.2)
Black/African American 134 (14.6) 17 (11.1) 44 (10.8)
White 688 (74.9) 98 (64.1) 280 (68.5)
Asian 55 (6.0) 5 (3.3) 35 (8.6)
Biracial/Interracial -- 17 (11.1) 14 (3.4)
Other 41 (4.5) -- 1 (0.2)
Ethnicity
Hispanic/Latinx 106 (11.5) 43 (28.1) 60 (14.7)
Not Hispanic/Latinx 918 (89.6) 104 (68.0) 324 (79.2)
Sleep Diary
SOL (minutes) 21.17 ± 20.15 -- --
WASO (minutes) 7.01 ± 10.41 -- --
TST (minutes) 447.20 ± 64.17 -- --
SE (%) 90.83 ± 5.98 -- --
PSQI 5.54 ± 2.74 -- --
MFI-GF 11.71 ± 3.49 -- --
DBAS-16 62.97 ± 23.34 -- --
ESS 8.68 ± 3.62 -- --
MEQ-7 15.93 ± 4.34 -- --
QIDS 6.62 ± 3.92 -- --
STAI 39.57 ± 10.41 -- --
PSS 18.22 ± 7.27 -- --
AUDIT 4.26 ± 4.71 -- --
MPS 0.97 ± 2.88 -- --
ISI 7.21 ± 5.04 8.76 (7.53) 7.70 ± 5.62
ISI Total Score ≥ 11
Insomnia Group 259 (25.1) 65 (42.5) 116 (29.4)
No insomnia Group 770 (74.6) 88 (57.5) 279 (70.6)
ISI Total Score ≥ 15
Insomnia Group 120 (11.6) 45 (29.4) 55 (13.9)
No insomnia Group 910 (88.2) 108 (70.6) 340 (86.1)
Insomnia Diagnosis
Insomnia -- 71 (46.4) --
No Insomnia -- 82 (53.6) --
Open in a new tab

SOL, sleep onset latency; WASO, wake time after sleep onset; TST, total sleep time; SE, sleep efficiency; PSQI, Pittsburgh Sleep Quality Index; MFI-GF, Multidimensional Fatigue Inventory-General Fatigue; DBAS-16, Dysfunctional Beliefs about Sleep Scale-16 Item; ESS, Epworth Sleepiness Scale; MEQ-7, Morningness-Eveningness Questionnaire-7 Item; QIDS, Quick Inventory of Depressive Symptoms; STAI, State-Trait Anxiety Inventory; PSS, Perceived Stress Scale; AUDIT, Alcohol Use Disorders Identification Test; MPS, Marijuana Problems Scale; ISI, Insomnia Severity Index.

Measures

Single-Time-Point Retrospective Self-Report Measures of Sleep.
Insomnia Severity Index (ISI):

The ISI is a self-report measure designed to assess current perceived insomnia symptom severity (e.g., over the past two weeks; Morin, 1993). The measure consists of 7 items evaluating the severity of problems with sleep-onset, sleep maintenance, early morning awakenings, satisfaction with current sleep pattern, interference with daily functioning, noticeability of the sleep problem to others, and the level of worry or distress caused by the sleep problem. Items are rated on a 5-point Likert scale with responses ranging from 0 to 4 (anchors vary). Total scores range from 0 to 28, with higher scores representing greater insomnia severity. Clinical cutoff scores of ≥ 11 (Bastien et al., 2001; Morin et al., 2011) and ≥ 15 (Morin, 1993; Morin et al., 1999) have been established to reflect an insomnia complaint above clinical threshold in adults. In the current study, the ISI yielded a Cronbach α of .86, compared to the value reported in the original validation study of .74 (Bastien et al., 2001).

Pittsburgh Sleep Quality Index (PSQI):

The PSQI is a self-report measure which assesses sleep quality and sleep disturbances, broadly defined (Buysse et al., 1989). The version used in the current study was a 19-item self-report retrospective questionnaire of the past 7 days designed to measure 7 domains or components: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleep medications, and daytime dysfunction. Component scores range from 0 to 3, and, when summed, produce a global score ranging from 0 to 21, with higher scores indicating more severe sleep difficulties; a cutoff of >5 is indicative of poor sleep quality. In the current study, the PSQI yielded a Cronbach α of .59.

Dysfunctional Beliefs about Sleep Scale-16 Item (DBAS-16):

The DBAS-16 is a 16-item self-report measure used to evaluate the role of sleep related beliefs and attitudes in insomnia (Morin et al., 2007). Statements reflecting one’s beliefs and attitudes about sleep are scored on a Likert scale 0 (strongly disagree) to 10 (strongly agree). Higher scores are associated with greater dysfunctional beliefs and negative impact on sleep quality. In the current study, the DBAS-16 yielded a Cronbach α of .85.

Epworth Sleepiness Scale (ESS):

The ESS is a self-report measure designed to assess daytime sleepiness (Johns, 1991). The measure consists of 8 items on a 4-point Likert scale on which respondents rate their likeliness of dozing in each situation, from 0 (would never doze) to 3 (high chance of dozing). Total scores range from 0 to 24, with higher scores representing greater sleepiness. Scores over 10 suggest significant daytime sleepiness, and scores over 15 suggest pathological sleepiness associated with conditions like sleep-related breathing disorders or narcolepsy. In the current study, the ESS yielded a Cronbach α of .68.

Multidimensional Fatigue Inventory (MFI):

The MFI is a 20-item self-report measure designed to assess fatigue across five dimensions: general fatigue, physical fatigue, reduced motivation, reduced activity, and mental fatigue (Smets et al., 1995). Each subscale consists of 4 items on a 5-point Likert scale ranging from 1 (yes, that is true) to 5 (no, that is not true). Total scores range from 4 to 20, with higher scores indicating greater levels of fatigue. The current sample yielded Cronbach α of .76 (general fatigue), .79 (physical fatigue), .57 (reduced motivation), .77 (reduced activity), and .86 (mental fatigue).

Daily Self-Report Measure of Sleep.
Sleep Diary:

Participants were asked to complete a paper-and-pencil sleep diary each morning upon awakening for 7 days. The sleep diary is a subjective, retrospective measure that asks participants estimate their sleep on the previous night (e.g., bedtime, sleep onset, night-time awakenings, morning wake time). These variables allow for the calculation of additional sleep parameters including sleep onset latency (SOL), wake after sleep onset (WASO), time in bed (TIB), total sleep time (TST), and sleep efficiency (SE). Sleep diaries are a reliable instrument in collecting sleep data (Aili et al., 2017; Dietch & Taylor, 2021; Rogers et al., 1993). The Consensus Sleep Diary (Carney et al., 2012) was not used because it had not been developed at the time this data was collected, but the version of the sleep diary used in the current study was similar in most ways to the Consensus Sleep Diary. For a detailed description of the sleep diary used in the current study, please see prior publication (Taylor et al., 2013).

Single-Time-Point Retrospective Self-Report Measures Psychosocial Symptoms.
Quick Inventory of Depressive Symptoms (QIDS):

The QIDS (Rush et al., 2003) a brief version of the 30-item Inventory of Depressive Symptomatology (Rush et al., 1996), is a self-report measure designed to assess depressive symptoms experiences during the past week. The measure consists of 16 items on a 4-point Likert scale (0 to 3). Total scores range from 0 to 48, with higher scores indicating greater depressive symptomatology. In the current study, the QIDS yielded a Cronbach α of .71.

State-Trait Anxiety Inventory (STAI):

The STAI is a self-report measure designed to assess general anxiety symptoms (Spielberger et al., 1983; Spielberger et al., 1970). The trait anxiety subscale consists of 20 statements presented on a 4-point scale. Total scores range from 20 to 80, with higher scores indicating greater anxiety symptomatology. In the current study, the STAI yielded a Cronbach α of .92.

Perceived Stress Scale (PSS):

The PSS is a self-report measure that assesses several domains of stress including unpredictability, lack of control, burden overload, and stressful life circumstances in the past month (Cohen et al., 1983). The measure consists of 14 items on a 5-point Likert scale ranging from 0 (never) to 4 (very often). Total scores range from 0 to 56, with higher scores representing greater perceived stress. In the current study, the PSS yielded a Cronbach α of .88.

Alcohol Use Disorders Identification Test (AUDIT):

The AUDIT is a self-report screening instrument used to detect alcohol consumption that has become harmful to health (Saunders et al., 1993). The measure consists of 10 items (yes/no and multiple choice) on a 5-point Likert scale ranging from 0 to 4. Total scores range from 0 to 40, with higher scores representing greater harmful alcohol use. A score of ≥ 8 indicates problematic alcohol use. In the current study, the AUDIT yielded a Cronbach α of .82.

Marijuana Problems Scale (MPS):

The MPS is an adapted self-report measure that assesses negative consequences of excessive marijuana in social, occupational, physical, and personal domains over the past 90 days (Stephens et al., 2000). The measure includes 20 items on a 3-point scale ranging from 0 (no problem) to 2 (serious problem). Total scores range from 0 to 40, with higher scores representing greater marijuana-related consequences. In the current study, the MPS yielded a Cronbach α of .90.

Procedure and Analyses

Participants in Sample I completed the above measures as part of a larger epidemiological study of health in college students. Following informed consent, participants were asked to complete a sleep diary for 7 days. At the end of the week, they returned the sleep diary to study personnel and then completed all other study questionnaires to ensure the questionnaires assessed a similar period to the sleep diaries.

The degree of agreement between the ISI items and total scores and other measures of sleep was assessed by calculating Pearson correlation coefficients between the ISI items and total score with weekly averages from the sleep diary and other sleep questionnaires (i.e., PSQI, MFI, DBAS-16, and ESS). Comparisons between the ISI items and total score with other psychosocial measures were assessed by calculating correlation coefficients between the ISI and the psychosocial questionnaires (i.e., QIDS, STAI, PSS, AUDIT, and MPS). Analyses were performed using SPSS 26 (2021).

Results

The ISI total score mean was 7.21 (SD = 5.04; Table 2). There were significant sex differences in ISI total score in this sample F1, 978= 9.57, p = .002, Cohen’s d =0.23 (Table 1), with females (M = 7.56) reporting significantly higher scores on the ISI than males (M = 6.42). Significant correlations (ps < .05) were demonstrated between the ISI total score and sleep diary parameters, sleep questionnaires, and psychosocial questionnaires.

Table 2.

Experiment I: Correlations between ISI items, sleep diary, and common sleep questionnaires

Sleep Diary Parameters Sleep Questionnaires
ISI Item Mean (SD) SOL WASO TST SE PSQI DBAS-16 ESS MFI-GF MFI-PF MFI-RA MFI-RM MFI-MF
1. Falling asleep 0.89 (0.99) .40 .28 −.12 −.44 .55 .35 .05 .34 .18 .21 .18 .23
2. Staying asleep 0.64 (0.88) .21 .45 −.08 −.36 .51 .32 .10 .30 .18 .18 .17 .18
3. Waking too early 0.74 (1.02) .12 .24 −.09 −.23 .32 .24 .12 .26 .13 .13 .15 .20
4. Satisfied 1.98 (1.01) .20 .19 −.20 −.29 .49 .39 .20 .57 .32 .29 .37 .35
5. Interference 1.42 (1.06) .14 .17 −.16 −.24 .43 .50 .28 .57 .30 .29 .36 .37
6. Noticeable 0.77 (0.95) .15 .15 −.11 −.22 .41 .46 .24 .48 .29 .27 .32 .34
7. Worry/Distress 0.77 (0.96) .19 .22 −.17 −.29 .48 .51 .25 .50 .28 .25 .33 .32
Total 7.21 (5.04) .27 .33 −.18 −.40 .62 .54 .25 .59 .33 .32 .37 .39
Open in a new tab

ISI, Insomnia Severity Index; SOL, sleep onset latency; WASO, wake time after sleep onset; TST, total sleep time; SE, sleep efficiency; PSQI, Pittsburgh Sleep Quality Index; DBAS-16, Dysfunctional Beliefs about Sleep Scale-16 Item; ESS, Epworth Sleepiness Scale; MEQ-7,; MFI-GF, Multidimensional Fatigue Inventory-General Fatigue; MFI-PF, Multidimensional Fatigue Inventory-Physical Fatigue; MFI-RA, Multidimensional Fatigue Inventory-Reduced Activity; MFI-RM, Multidimensional Fatigue Inventory-Reduced Motivation; MFI-MF, Multidimensional Fatigue Inventory-Mental Fatigue. All bolded correlation pairs were significant at p < .01

The ISI total score demonstrated moderate (i.e., .3 < r ≤ .6) to strong (i.e., .6 < r ≤ .9) correlations with the DBAS-16 (r = .54), MFI-GF (r = .59), and PSQI (r = .62) and weak correlations with the MFI-MF (r = .39), MFI-RM (r = .37), MFI-PF (r = .33), MFI-RA (r = .32), and ESS (r = .25). Weak (i.e., .0 < r ≤ .3) to moderate correlations were found between the ISI total score and sleep diary parameters (Table 2: rs = .18-.40). The ISI items demonstrated moderate (approaching strong) correlations with corresponding sleep diary parameters (i.e., sleep onset latency r = .40, wake after sleep onset r = .45).

The ISI total score showed weak, but significant correlations with the AUDIT (r = −.10) and the MPS (r = .11), suggesting minimal overlap with insomnia severity and alcohol abuse or substance abuse. The ISI total score showed moderate correlations with the QIDS (r = .57), STAI (r = .49) and PSS (r = .46), suggesting some overlap of insomnia severity with symptoms of depression, anxiety, and stress.

Experiment II

The aim of Experiment II was to extend the evaluation of the psychometric properties of the ISI to assess its diagnostic accuracy as an insomnia screening tool in a clinical sample of undergraduates with and without insomnia disorder.

Methods

Sample II

This secondary analysis utilized previously collected data (2011-2012). A total of 153 healthy college students (71 with insomnia, 82 normal sleepers) volunteered to participate in the parent study in exchange for financial incentives totaling $105. The sample included undergraduate students with an average age of 20.24 years (SD = 2.57) and 57.5% were females (2.6% did not provide sex information). Participants in the parent study were required to be 18-29 years old and enrolled at the University. Participants were excluded for having a sleep disorder other than insomnia, an Axis I or II psychiatric disorder (e.g., mental health disorder, substance use disorders, personality disorders, intellectual disability), serious chronic medical condition (e.g., cancer, HIV, pain, immune, Guillain-Barre syndrome), and could not be pregnant or nursing. Sample characteristics are presented in Table 1. The racial and ethnic diversity was similar to Sample I. This study was approved by the university institutional review board, and informed consent was obtained for all participants.

Measures

ISI:

The ISI (described above) used in Experiment II was a 7-item retrospective questionnaire of the past 2 weeks. Cronbach α in Sample II was .94.

Structured Clinical Interview Schedule for DSM-5 Sleep Disorders:

Participants were interviewed by trained master’s level psychology student interviewers using a Structured Clinical Interview Schedule to assess for current sleep disorders (Taylor et al., 2019) as defined by the Diagnostic and Statistical Manual, fifth edition (DSM-5; (American Psychiatric Association, 2013). Co-author (DJT), a licensed psychologist certified in behavioral sleep medicine used the existing criteria of the DSM-IV-TR in conjunction with the proposed changes for the DSM-5 to create individual interview questions for all the sleep-wake disorder categories. Of the proposed changes, only those that were ultimately included in the DSM-5 are in the final version of the SCISD.

Interviewers were supervised by a licensed clinical psychologist board certified in both Sleep Medicine and Behavioral Sleep Medicine. For the purposes of this study, only participants who were classified as either meeting criteria for insomnia disorder or not meeting criteria for any sleep disorder (“normal sleepers”) were included for analyses.

Procedure and Analyses

Participants in Sample II completed a series of questionnaires, including the ISI, and partook in structured clinical interviews for sleep disorders. The ability of the ISI to differentiate between individuals with and without insomnia was calculated using analyses of sensitivity, specificity, the receiver operating characteristic (ROC) curve, and the Youden Index. These analyses were performed using DSM-5 insomnia diagnosis as the “gold standard” against which the ISI was compared. DSM-5 diagnostic criteria for insomnia disorder includes complaints of difficulties with falling asleep, staying asleep, or waking too early, causing significant distress or impairment in general functioning, occurring at least 3 nights per week for at least three months.

Sensitivity and specificity were calculated with True and False Negatives and Positives. True Negatives (TN) were classified as being negative for both insomnia diagnosis (DSM-5) and ISI criteria. True Positives (TP) were classified as being positive for both insomnia disorder and ISI criteria. False Negatives (FN) were classified as positive for insomnia disorder, but negative for ISI criteria. Finally, False Positives (FP) were classified as negative for insomnia disorder, but positive for ISI criteria.

Sensitivity, the probability the ISI can correctly detect individuals with insomnia, was calculated by TP / (TP + FN). Specificity, the probability the ISI can correctly identify individuals without insomnia, was calculated by TN / (TN + FP). The ROC curve plots sensitivity (y-axis) and 1 – specificity (x-axis) which produces the area under the curve (AUC) summary measure to help assess diagnostic accuracy (Zweig & Campbell, 1993). The Youden Index is a single statistic of the ROC curve used as a criterion for selecting the optimum cutoff point when a diagnostic test gives a numeric result by optimizing the test’s differentiating ability when utilizing both sensitivity and specificity (Youden, 1950). Analyses were performed using SPSS 26 (2021).

Results

People with insomnia (n = 71) reported significantly higher average ISI scores (M = 16.08, SD = 4.06) than those without (n = 82; M = 2.54, SD = 2.19), t (100.68) = 25.51, p <.001.

Table 4 presents the sensitivity and specificity of the ISI and Youden Index scores at various cutoff points compared to the “gold standard” of DSM-5 insomnia disorder diagnosis. The AUC was .999, suggesting the ISI has extremely high sensitivity and specificity in its detection of insomnia diagnosis within this sample. The Youden Index score was maximized (J = .96; sensitivity = .97 and specificity = .99) at a cutoff score of ≥ 8, which is lower than the previously suggested clinical cutoff scores of 11 (Bastien et al., 2001; Morin et al., 2011) and 15 (Morin, 1993; Morin et al., 1999). However, specificity was maximized (1.00) at ≥ 10, which may maximize clinical utility for detecting insomnia in college students with this measure.

Table 4.

Experiment III: Confirmatory Factor Analyses - Fit indices for the estimated models

Sampled I (N = 1,040)
Model χ2 p df RMSEA CFI TLI SRMR
One-factor 91.469 .00 14 0.074 0.978 0.967 0.069
Two-factor 20.621 .08 13 0.024 0.998 0.997 0.031
Three-factor 9.386 .40 9 0.006 1.000 1.000 0.021
Sample III (N= 395)
Model χ2 p df RMSEA CFI TLI SRMR
One-factor 49.82 .00 14 0.081 0.980 0.970 0.078
Two-factor 16.69 .21 13 0.027 0.998 0.997 0.044
Three-factor 3.67 .93 9 0.000 1.000 1.007 0.021
Open in a new tab

RMSEA, root mean square error of approximation; CFI, confirmatory factor index; TLI, Tucker Lewis Index; SRMR; Standardized Root Mean Square Residual

Experiment III

The aim of Experiment III was to extend the evaluation of the psychometric properties of the ISI to assess the test-retest reliability of the ISI (Sample III) and to conduct a confirmatory factor analysis (CFA) and Item Response Theory via Graded Response Model (GRM) of the ISI in two longitudinal designed studies (Samples I and III).

Methods

Sample I

For details about Sample I, see “Experiment I” above.

Sample III

This secondary analysis utilized previously collected data (Spring 2014-Spring 2015). College students were recruited from a large public university. A total of 409 participants completed an online questionnaire battery via REDCap electronic data capture tools (Harris et al., 2019; Harris et al., 2009) and completed the same battery 30-35 days later. Participants did not receive any intervention between these two time points. This study was approved by the university institutional review board, and informed consent was obtained for all participants. Sample characteristics are presented in Table 1.

The final sample for this analysis consisted of 395 college students with an average age of years 20.79 (SD = 3.44) and 69.9% were females. The racial and ethnic diversity was similar to Studies I and II. Age, race, and ethnicity were not significantly different in the subsets used in the test-retest analysis (n = 121) or the confirmatory factor analysis (n = 395), described below. For the test-retest analysis, participants who did not complete the ISI at both time points were excluded. This subset consisted of 121 college students and was predominately female (81.0%).

Measures

ISI:

The ISI (described above) used in Experiment III was a 7-item retrospective questionnaire of the past 2 weeks. Cronbach α in Sample III was .86 (time 1) and .90 (time 2).

Procedure and Analyses

Participants in Sample III completed a series of questionnaires, including the ISI, at two time points approximately one month apart. The average ISI total score for the sample at time 1 was M = 8.00 (SD = 5.42). Test-retest reliability of the ISI was assessed by the intraclass correlation coefficient (ICC). ICC estimate and its 95% confident intervals were calculated using SPSS 26 (2021) based on a mean-rating (k = 2), absolute-agreement, two-way mixed-effects model (Koo & Li, 2016).

Confirmatory Factor Analysis.

Confirmatory factor analysis (CFA) was used to test alternative models to confirm the factor structure of the ISI in Samples I and III. The ISI was tested utilizing three distinct CFA models based on factor structures generated in prior exploratory factor analyses (Bastien et al., 2001; Fernandez-Mendoza et al., 2012). A one-factor model included all items onto “Insomnia.” A two-factor model included Factor 1 (Symptoms): initial, middle, and terminal insomnia (items 1, 2, 3) and Factor 2 (Distress): satisfaction, interference, noticeability, and distress (items 4, 5, 6, 7). A three-factor model included Factor I (Impact): interference, noticeability, distress (items 5, 6, 7), Factor II (Severity): initial, middle, terminal (items 1, 2, 3), and Factor III (Satisfaction): initial, satisfaction, distress (items 1, 4, 7).

Assumptions of normality for all items was violated (Shapiro-Wilk’s p < .05). Thus, diagonally weighted least squares estimator was used (Mindrila, 2010). Considering the various model fit indices, the three-factor model had the lowest value of chi-square (χ2 = 9.386; χ2III = 3.67) and root mean square error of approximation (RMSEA) (RMSEAI = 0.006; RMSEAIII = 0.000), the comparative fit index (CFII = 1.000; CFIIII = 1.000) and Tucker-Lewis index (TLII; 1.000; TLIIII; 1.007) values were highest (Bentler, 1990). Generally, RMSEA smaller than .06 and a CFI and TLI larger than .95 indicate relatively good model–data fit (Hu & Bentler, 1999). The CFA was performed using R version 4.3.1 (R Core Team) and the lavaan package version 0.6-7.

Item Response Theory.

Samejima’s (1969, 1997) Graded Response Model, Muraki’s (Muraki, 1992; van der Linden & Hambleton, 2013) Generalized Partial Credit Model, and Bock’s (1972) Nominal Response Models are extensions of the two-parameter logistic model in Item Response Theory (IRT), allowing for the application of IRT on polytomous responses (i.e., Likert-scales). IRT offers specific models for examining and organizing items in instruments and an evaluative process that cannot be conducted with classical test theory or exploratory/confirmatory factor methods.

For Samples I and III, a unidimensional IRT model was implemented, only latent trait (Θ) was considered for insomnia severity. The model calculated, for each item, a slope parameter (a) and 4 threshold parameters (bm). Higher parameters (a) reflected a greater strength of the relationship between each item on the ISI an overall insomnia severity. Lastly, the item information function (IIF) was computed for each item to indicate reliability that each item could provide to overall insomnia severity and the reliability of the ISI was assessed in IRT framework with the test information function (TIF). The GRM analyses were conducted in IRTPRO, Version 6.0 (Vector Psychometric Group, n.d.).

Results

The ISI showed good test-retest reliability (Koo & Li, 2016), with the total score ICC = .87 (95% C.I. .81 - .91), F = 7.78, p <.001, and item-level ICC ranging from .67-.87 (allps <.001). Test-retest reliability for the ISI total and items and inter-item and item-total correlations for Sample III are presented in the Supplemental Materials Table S2 and S3, respectively.

CFA (Table 4) showed that a three-factor model provided a better fit to the data than a one- or two-factor model for both Samples I and III. Pattern coefficients for the one-factor (insomnia), two-factor (symptoms, distress), and three-factor (impact, severity, satisfaction) models are presented in Table 5.

Table 5.

Experiment III: Confirmatory Factor Analyses - Standardized pattern coefficients

Sample I (N = 1,040)
One-Factor Two-Factor Three-Factor
ISI Item Insomnia (λ) Symptoms (λ) Distress (λ) Impact (λ) Severity (λ) Satisfaction (λ)
1. Falling asleep 0.59 0.73 0.61 0.12
2. Staying asleep 0.58 0.73 0.76
3. Waking too early 0.43 0.51 0.53
4. Satisfied 0.75 0.77 0.77
5. Interference 0.80 0.82 0.88
6. Noticeable 0.74 0.77 0.81
7. Worry/Distress 0.85 0.86 0.19 0.68
Sample III (N= 395)
One-Factor Two-Factor Three-Factor
ISI Item Insomnia (λ) Symptoms (λ) Distress (λ) Impact (λ) Severity (λ) Satisfaction (λ)
1. Falling asleep 0.67 0.78 0.45 0.31
2. Staying asleep 0.70 0.84 0.92
3. Waking too early 0.54 0.62 0.66
4. Satisfied 0.74 0.76 0.77
5. Interference 0.80 0.85 0.95
6. Noticeable 0.65 0.68 0.75
7. Worry/Distress 0.83 0.86 0.16 0.71
Open in a new tab

ISI, Insomnia Severity Index

The individual assessment of each item to the GRM indicated that Items 5 (i.e., interference) and 6 (i.e., noticeability) in Sample I and Item 7 (i.e., worry/distress) in Sample III did not have a good fit for the 2PL model (p < .01). For full item-level results, see Supplemental Files Table S5. Table 6 shows the results of the item parameters of the GRM. For Sample I the items span a wide portion of insomnia severity, from −1.85 (b1 for Item 4: Satisfied) to 5.00 (b4 for Item 3: Waking Too Early). Similarly, for Sample III the items span from −1.57 (b1 for Item 4: Satisfied) to 3.67 (b4 for Item 3: Waking Too Early).

Table 6.

Experiment III: Graded response model parameter estimates for the ISI, a(Θ-b)

Sample I (N = 1,040)
ISI Item a (s.e.) b1 (s.e.) b2 (s.e.) b3 (s.e.) b4 (s.e.)
1. Falling asleep 1.32 (0.09) −0.25 (0.06) 0.99 (0.08) 2.61 (0.17) 3.47 (0.25)
2. Staying asleep 1.32 (0.10) 0.30 (0.06) 1.47 (0.10) 3.00 (0.21) 4.50 (0.40)
3. Waking too early 0.89 (0.08) 0.35 (0.08) 1.55 (0.14) 3.19 (0.28) 5.00 (0.48)
4. Satisfied 2.47 (0.14) −1.85 (0.09) −0.53 (0.05) 0.58 (0.05) 1.94 (0.09)
5. Interference 3.16 (0.19) −0.92 (0.05) 0.20 (0.04) 1.12 (0.06) 2.15 (0.10)
6. Noticeable 2.88 (0.18) 0.04 (0.04) 0.95 (0.05) 1.77 (0.08) 2.81 (0.16)
7. Worry/Distress 4.19 (0.31) 0.04 (0.04) 0.90 (0.05) 1.64 (0.07) 2.38 (0.11)
Sample III (N= 395)
ISI Item a (s.e.) b1 (s.e.) b2 (s.e.) b3 (s.e.) b4 (s.e.)
1. Falling asleep 1.76 (0.17) −0.28 (0.09) 0.61 (0.09) 1.75 (0.15) 2.71 (0.26)
2. Staying asleep 1.92 (0.20) −0.01 (0.08) 0.92 (0.10) 1.97 (0.17) 3.57 (0.46)
3. Waking too early 1.28 (0.15) −0.13 (0.10) 1.22 (0.14) 2.55 (0.27) 3.67 (0.45)
4. Satisfied 2.50 (0.22) −1.57 (0.12) −0.51 (0.08) 0.42 (0.08) 1.60 (0.12)
5. Interference 3.27 (0.32) −0.62 (0.08) 0.45 (0.07) 1.22 (0.09) 1.88 (0.13)
6. Noticeable 2.23 (0.23) 0.10 (0.08) 0.98 (0.09) 1.73 (0.14) 2.56 (0.23)
7. Worry/Distress 3.98 (0.44) −0.04 (0.07) 0.74 (0.07) 1.40 (0.09) 1.87 (0.13)
Open in a new tab

ISI, Insomnia Severity Index; a, slope parameter; b1,…b4, threshold parameters; s.e., standard error

The values derived from the IIF and TIF for levels of insomnia severity are presented in Table 7. In Sample I, the results are more informative for middle- to high-levels of insomnia severity (0 ≤ Θ ≤ 2.4), where the standard error in the measurement is smaller than at low or high levels. Sample III the results are more informative for middle the middle levels of insomnia severity (0 ≤ Θ ≤ 1.6). Goodness of fit statistics calculated based on Log Likelihood for Sample I: 15,075.39 (−2loglikelihood); 15,145.39 (AIC); 15,318.37 (BIC) and Sample III: 6,040.81 (−2loglikelihood); 6,110.81 (AIC); 6,250.07 (BIC). Item-level graphical representations of the GRM for Sample I and III are provided in the Supplemental Files.

Table 7.

Experiment III: Graded response model: Item information function and Test information function

Sample I (N = 1,040) Θ
ISI Item −2.4 −1.6 −0.8 0 0.8 1.6 2.4
1. Falling asleep 0.09 0.22 0.40 0.51 0.52 0.51 0.52
2. Staying asleep 0.05 0.12 0.27 0.44 0.52 0.52 0.51
3. Waking too early 0.06 0.10 0.16 0.21 0.24 0.24 0.24
4. Satisfied 1.00 1.54 1.55 1.58 1.57 1.48 1.13
5. Interference 0.09 0.93 2.55 2.51 2.54 2.29 2.17
6. Noticeable 0.01 0.07 0.62 2.17 2.30 2.33 2.06
7. Worry/Distress 0.00 0.02 0.49 4.43 4.51 4.72 4.52
Test 2.29 4.00 7.03 12.85 13.21 13.09 12.15
Expected s.e. 0.66 0.50 0.38 0.28 0.28 0.28 0.29
Sample III (N= 395) Θ
ISI Item −2.4 −1.6 −0.8 0 0.8 1.6 2.4
1. Falling asleep 0.07 0.25 0.65 0.91 0.92 0.92 0.91
2. Staying asleep 0.04 0.16 0.54 1.02 1.10 1.07 0.95
3. Waking too early 0.08 0.19 0.35 0.47 0.49 0.50 0.50
4. Satisfied 0.62 1.64 1.69 1.74 1.60 1.64 0.66
5. Interference 0.03 0.40 2.48 2.33 2.88 3.03 1.41
6. Noticeable 0.02 0.11 0.52 1.30 1.48 1.51 1.43
7. Worry/Distress 0.0 0.03 0.70 4.14 4.34 4.64 1.54
Test 1.86 3.78 7.94 12.91 13.82 14.30 8.39
Expected s.e. 0.85 0.51 0.35 0.28 0.27 0.26 0.35
Open in a new tab

ISI, Insomnia Severity Index; s.e., standard error

Note: Marginal Reliability for Response Pattern Scores: 0.88 (Sample I) and 0.89 (Sample III)

Discussion

The ISI showed strengths among college students and performed well as a screening tool for insomnia severity in college students. Results indicated the ISI serves as an accurate diagnostic screening tool for insomnia in this population (healthy college students with no comorbidities). The ISI demonstrated good psychometric properties and appears generally valid for screening insomnia disorder and assessing insomnia severity in college students. However, overlap with psychological symptoms suggests caution while interpreting these constructs independently.

The results of Experiment I demonstrated the ISI was strongly related to global sleep quality (PSQI), general fatigue (MFI-GF), and dysfunctional beliefs about sleep (DBAS-16). The ISI total score was moderately related to the analogous sleep diary sleep efficiency (r = −.40), as well as most of those assumed to be analogous (i.e., Item 1/SOL; Item 2/WASO), again with moderate correlations. One reason correlations may not have been higher is that the ISI is an assessment over the course of 1-2 weeks whereas the values of the daily sleep diary are provided daily for the previous night’s sleep. Thus, the participants might focus on the worse nights of the past 1-2 weeks (i.e., negativity bias, disqualifying the positive, magnification of the negative) when completing the ISI, but those more extreme nights are weighed down by the more frequent better nights on the sleep diary.

The correlation of the ISI total score with the measures of general fatigue (Morin et al., 2011), global sleep quality (Kaufmann et al., 2019; Morin et al., 2011; Veqar & Hussain, 2017), and sleep diary derived variables (i.e., SOL, WASO; Wong et al., 2017) were similar to previous studies. However, the correlations between ISI and global sleep quality varied between previous studies (rs = .45-.80; (Kaufmann et al., 2019; Morin et al., 2011; Veqar & Hussain, 2017). Additionally, the correlation between the ISI and WASO was weaker in the current sample. The association between sleep diary derived sleep efficiency and the ISI total score was nearly twice the size in our sample compared to previous studies (Bastien et al., 2001; Wong et al., 2017). This may occur due to differences in the measures themselves, with daily (sleep diary) vs. weekly (ISI) retrospective data being assessed, resulting in increased accuracy in daily (compared to weekly) estimate. Another possibility could be due to sample differences. One of these studies was assessed in a sample with ages that span a much larger range than that of the current study (Bastien et al., 2001). Both studies were a fraction of the current sample size (Ns = 145 and 158 vs. 1,040). Additionally, racial/ethnic/cultural differences may also exist between students at an American vs. East Asian University resulting in these differences (Wong et al., 2017). Comparisons of SOL and WASO with analogous items on the ISI were similar to previous findings with slightly stronger correlations in the current sample (Bastien et al., 2001).

Although the correlations of the ISI with analogous sleep diary variables were significant, the magnitudes were moderate at best. This is possibly because the ISI classifies sleep onset and maintenance severity scored on a Likert-scale, rather than a time estimation (i.e., minutes) used within the sleep diary. So, while related, these two assessments may be measuring distinct components of SOL and WASO. Additionally, the severity rating of SOL and WASO could be related to one’s self-ascribed perception of their sleep instead of the duration of wakefulness (e.g., insomnia identity, the conviction that one has insomnia, a sleep complaint that can be measured independently of sleep; (Lichstein, 2017). This is further conceptualized given that not all who reported insomnia indicated insomnia-related impairments on the sleep diary (e.g., SOL and/or WASO less than 30 minutes; (Emert et al., 2021). Lastly, this result may be due to the single-time-point retrospective estimate of the ISI, which has known issues of bias such as systematic overestimation (Johnson et al., 2011).

The ESS demonstrated statistically significant, but weak correlations with the ISI total score and ISI items. These results are consistent with previous research showing that insomnia, which is principally what the ISI is measuring, is highly correlated with daytime fatigue but not with excessive daytime sleepiness (e.g., Riedel & Lichstein, 2000). It appears the ISI is slightly less sensitive to conditions that produce excess daytime sleepiness compared to global sleep quality or fatigue.

The ISI item and total scores overlapped with depression (QIDS), anxiety (STAI), and perceived stress (PSS) total scores; high overlap between these scores indicate the potential for some shared symptomatology (e.g., distress). Within this sample, the ISI was very weakly associated with alcohol abuse (AUDIT) and marijuana abuse (MPS).

The correlation of the ISI with the measure of anxiety (STAI) was nearly identical to previous research (Morin et al., 2011) and was similar, but weaker in the current sample, to the measure of sleepiness (Chung et al., 2011; Kaufmann et al., 2019). Though different assessment measures were used, these results were also consistent with the associations found when the ISI was compared to anxiety measured with the Beck Anxiety Inventory (r = .45, p ≤ .001) and depression measured with the Beck Depression Inventory (r = .48, p ≤ .001; Kaufmann et al., 2019).

The significant overlap between the ISI and the QIDS, STAI, or PSS is unsurprising given previous findings of moderate to strong correlations between the ISI and psychological symptoms (e.g., depression, anxiety, and perceived stress; (Morin et al., 2011). The QIDS has several questions that assess sleep, but even after these items were removed, correlations remained statistically significant and in the moderate range for both item-level and total ISI. Additionally, a recent descriptive study quantifying the repetition of symptoms across distinct disorders shows considerable overlap (Forbes et al., 2023). Insomnia symptoms, sleepiness, fatigue, difficulty concentrating, and irritability were symptoms in up to 22 other diagnoses (Forbes et al., 2023). While his overlap makes it difficult to understand these constructs independently, it may help to better understand diagnostic comorbities, transdiagnostic processes, and personalized evaluation and treatment by utilizing a symptom-level or symptom patterns vs. diagnostic-level approach.

The results of Experiment II demonstrated the ISI very accurately identified college students with and without a diagnosis of Insomnia Disorder per DSM-5 criteria. The average ISI total score for individuals with insomnia disorder (M = 16.07, SD = 4.01) was similar to the original ISI validation study (see above). This suggests the ISI performed similarly in this population compared to the mostly middle-aged adult sample (Mage = 41.4, SD = 13.1, range = 17-82), examined in the original validation study (Bastien et al., 2001). The average ISI total score for the sample in Experiment I was M = 7.21 (SD = 5.04). Unsurprisingly, this was lower than the original validation study which only included participants with an insomnia complaint (M = 19.7, SD = 4.10; (Bastien et al., 2001). The average ISI total score for the sample in Experiment III at time one was M = 7.70 (SD = 5.62), which is again lower than in the original validation study, likely due to the mixed sample (i.e., not specific to patients with insomnia complaint or disorder).

The ISI performed somewhat better at identifying DSM-5 diagnosed insomnia in this U.S. college population than it did in a previous sample of Chinese college students (Wong et al., 2017). Both study samples included similar samples sizes, similar group sizes for insomnia and controls, and were screened for medical comorbidities and included healthy participants. It is possible that the current study performed a more extensive assessment (e.g., self-report measures, clinical interviews, history and physicals, lab assessments) to select otherwise healthy people (e.g., no other underlying sleep, psychiatric, or medical disorders) with and without insomnia, as screening procedures were not detailed in Wong et al. (2017). Another possible source of difference is the population and the version of the ISI used for assessing insomnia severity in each study (i.e., English vs. Chinese; Chung et al., 2011; Morin, 1993), which may alter comparative generalizability or psychometrics.

The optimal sensitivity and specificity for insomnia detection was a cutoff score of ≥ 8, providing confirmatory support for this cutoff value in a similarly aged-sample (Chung et al., 2011). This cutoff could be adjusted depending on the intended use of the ISI, as sensitivity and specificity are still very high at surrounding cutoff scores. Clinicians and researchers can examine sensitivity/specificity and choose the cutoff that achieves the balance between these values that is appropriate to their intended goals. This is particularly relevant within clinical contexts, since previous findings have established higher clinical cutoff scores ≥ 11 (Bastien et al., 2001; Morin et al., 2011) and ≥ 15 (Morin, 1993; Morin et al., 1999), which may indicate fewer college (or college-aged) students are screening positive for insomnia disorder using the ISI (i.e., False Negatives).

The results of Experiment III provided evidence for the test-retest reliability and factor structure of the ISI. These results from Samples I and III provided additional confirmatory support for a three-factor ISI model which includes the daytime impact of insomnia (impact), severity of sleep difficulties (severity), and worry about/dissatisfaction with sleep (satisfaction; Bastien et al., 2001; Fernandez-Mendoza et al., 2012). Additionally, this three-factor model is representative of the key criteria for insomnia disorder within the DSM-5.

The results of Experiment III provided evidence for the test-retest reliability, item-related information for insomnia severity as a unidimensional factor (GRM) and a possible factor structure (CFA) of the ISI. Test-retest reliability of the ISI, item-level and total score, were similar to previous studies of adolescents (Chung et al., 2011), University students (Veqar & Hussain, 2017), and adults (Lin et al., 2020). Item total correlations of the ISI were consistent with a previous sample of adolescents (Chung et al., 2011), but were stronger than those in a sample of adults (Lin et al., 2020). The cut-off value (≥ 8) estimated by maximization of sensitivity and specificity was directly between that found in two samples of adolescents and young adults. Previous cut-off values were ≥ 7 (sample aged 17-24 years; Chung et al., 2011; Wong et al., 2017) and ≥ 9 (sample aged 12-19 years; Chung et al., 2011). These results provided additional confirmatory support for a three-factor ISI model which includes the daytime impact of insomnia, severity of sleep difficulties, and sleep (dis)satisfaction (Bastien et al., 2001; Fernandez-Mendoza et al., 2012). Within this context, the ISI can be used to assess total severity of insomnia but also to assess nighttime sleep difficulties, sleep satisfaction, and the impact of insomnia on daytime functioning separately.

The results of the GRM for Sample I and III indicate that the ISI has an overall good fit to the model. Item level statistics (i.e., S-χ2) are non-significant (p > .01) for items 1-4 and 7 for Sample I and items 1-6 for Sample III. In Sample 1, items 5 and 6 indicated poor model fit (ps < .001) and in Sample III, item 7 indicated poor model fit (p = .007). For all S-χ2 item level results, see Supplemental File Table S5).

Based on results from the information function curves in Samples I and III, most items appear to assess insomnia severity better at values in the middle range (i.e., −1.0 < Θ < 3.0), with the exception of Item 4. In both samples, Item 3 does not appear to provide much information related to insomnia severity ability (rangeI = 0.04-0.25 and rangeIII = 0.05-0.50 spanning −2.8 ≤ Θ ≤ 2.8). Overall, the ISI is better at assessing individuals with moderate levels of insomnia severity (−0.40 ≤ Θ ≤ 2.0) and is not as good in assessing those with lower levels (Θ < −0.40) or higher levels of insomnia severity within this sample (Θ > 2.0; see Table 7).

The results of these studies are limited by several factors. First, the samples included for these studies were from a single institution and may not be generalizable to students at institutions in other regions or with other racial/ethnic and/or socioeconomic make-ups. Second, the sample used in Experiment II was healthy and without comorbidities, which limits generalizability to individuals with comorbid sleep, psychological, and medical conditions. Sample II was also rather small. The diagnostic accuracy of the ISI may also be altered among individuals with comorbidities. Future research should examine or develop measurement tools that account for college student’s unique sleep habits and should examine the psychometrics of the ISI in populations with insomnia disorder, and comorbid sleep and psychological conditions. Future research should also include comparisons between sensitivity to change evaluations vs. test-retest reliability to better assess the stability of insomnia severity over time, with and without intervention.

Though the ISI serves as an accurate diagnostic screening tool for insomnia in college students, the ISI should not be used alone to diagnose insomnia, and it should not be considered a substitute for a comprehensive sleep evaluation performed by an experienced sleep clinician including clinical interview, which would be optimal to accurately assign sleep disorder diagnoses. For instance, results from this study indicated the ISI is moderately sensitive to complaints of global sleep quality, fatigue, and subjective distress (i.e., depression, anxiety, and perceived stress), and is less sensitive to daytime sleepiness complaints, circadian preference, or alcohol or marijuana substance use disorders. Because the ISI may not accurately reflect impacts from daytime sleepiness, circadian preference, or substance abuse problems, assessment of these constructs may be warranted to better disentangle symptoms of insomnia from symptoms of depression, anxiety, and stress. A clinical interview may also better disentangle any symptom overlap, effects that negative affect has on subjective sleep distress report, and provide information related to other comorbidities that may affect diagnoses and treatment.

Supplementary Material

1
2
3
4

Table 3.

Experiment II: Sensitivity, specificity, and Youden’s Index of ISI compared to DSM-5 insomnia disorder diagnosis

ISI Cutoff Sensitivity Specificity Youden’s Index (J)
≥ 7 1.00 .95 .95
≥ 8 .97 .99 .96
≥ 10 .94 1.00 .94
≥ 11 .92 1.00 .92
≥ 15 .63 1.00 .63
Open in a new tab

ISI, Insomnia Severity Index

Highlights.

  • The Insomnia Severity Index was a valid and reliable measure in college students

  • The Insomnia Severity Index relates to sleep quality, fatigue, and sleep beliefs

  • The Insomnia Severity Index was accurate in the diagnosis of insomnia disorder

  • Results show evidence for test-retest reliability of the Insomnia Severity Index

  • Results show evidence for a three-factor structure of the Insomnia Severity Index

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References:

  1. Aili K, Åström-Paulsson S, Stoetzer U, Svartengren M, & Hillert L (2017). Reliability of actigraphy and subjective sleep measurements in adults: the design of sleep assessments. Journal of Clinical Sleep Medicine, 13(1), 39–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Albasheer OB, Al Bahhawi T, Ryani MA, Arishi AM, Hakami OM, Maashi SM, Al-Khairat HK, Alganmy OM, Sahal YA, & Sharif AA (2020). Prevalence of insomnia and relationship with depression, anxiety and stress among Jazan University students: A cross-sectional study. Cogent Psychology, 7(1), 1789424. [Google Scholar]
  3. American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). American Psychiatric Association. [Google Scholar]
  4. Bastien CH, Vallieres A, & Morin CM (2001). Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Medicine, 2, 297–307. 10.1016/S1389-9457(00)00065-4 [DOI] [PubMed] [Google Scholar]
  5. Bentler PM (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238. [DOI] [PubMed] [Google Scholar]
  6. Buysse DJ, Reynolds CF III, Monk TH, Berman SR, & Kupfer DJ (1989). The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Research, 28(2), 193–213. [DOI] [PubMed] [Google Scholar]
  7. Carney CE, Buysse DJ, Ancoli-Israel S, Edinger JD, Krystal AD, Lichstein KL, & Morin CM (2012). The consensus sleep diary: standardizing prospective sleep self-monitoring. Sleep, 35(2), 287–302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chahoud M, Chahine R, Salameh P, & Sauleau E (2017). Reliability, factor analysis and internal consistency calculation of the Insomnia Severity Index (ISI) in French and in English among Lebanese adolescents. eNeurologicalSci, 7, 9–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chung KF, Kan KKK, & Yeung WF (2011). Assessing insomnia in adolescents: comparison of insomnia severity index, Athens insomnia scale and sleep quality index. Sleep Medicine, 12(5), 463–470. [DOI] [PubMed] [Google Scholar]
  10. Cohen S, Kamarck T, & Mermelstein R (1983). A global measure of perceived stress. Journal of Health and Social Behavior, 385–396. [PubMed] [Google Scholar]
  11. Commerce, U. S. D. o. (2022). Recent high school completers and their enrollment in college, by sex and level of institution: 1960 through 2021. Census Bureau. https://nces.ed.gov/programs/digest/d22/tables/dt22_302.10.asp [Google Scholar]
  12. IBM Corp. (2021). (Version SPSS 26) IBM Corp. [Google Scholar]
  13. Darrell Bock R. (1972). Estimating item parameters and latent ability when responses are scored in two or more nominal categories. Psychometrika, 37(1), 29–51. [Google Scholar]
  14. Dietch JR, & Taylor DJ (2021). Evaluation of the Consensus Sleep Diary in a community sample: comparison with single-channel EEG, actigraphy, and retrospective questionnaire. Journal of Clinical Sleep Medicine, 17(7), 1389–1399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Dragioti E, Wiklund T, Alföldi P, & Gerdle B (2015). The Swedish version of the Insomnia Severity Index: Factor structure analysis and psychometric properties in chronic pain patients. Scandinavian Journal of Pain, 9, 22–27. [DOI] [PubMed] [Google Scholar]
  16. Emert SE, Gunn HE, Molzof HE, Dietch JR, & Lichstein KL (2021). Appraisals of insomnia identity in a clinical sample. Behaviour Research and Therapy, 145, 103943. [DOI] [PubMed] [Google Scholar]
  17. Fernandez-Mendoza J, Rodriguez-Muñoz A, Vela-Bueno A, Olavarrieta-Bernardino S, Calhoun SL, Bixler EO, & Vgontzas AN (2012). The Spanish version of the Insomnia Severity Index: a confirmatory factor analysis. Sleep Medicine, 13(2), 207–210. [DOI] [PubMed] [Google Scholar]
  18. Forbes MK, Neo B, Nezami OM, Fried EI, Faure K, Michelsen B, Twose M, & Dras M (2023). Elemental psychopathology: Distilling constituent symptoms and patterns of repetition in the diagnostic criteria of the DSM-5. Psychological Medicine, 1–9. doi: 10.1017/S0033291723002544 [DOI] [PubMed] [Google Scholar]
  19. Gagnon C, Bélanger L, Ivers H, & Morin CM (2013). Validation of the Insomnia Severity Index in primary care. The Journal of the American Board of Family Medicine, 26(6), 701–710. [DOI] [PubMed] [Google Scholar]
  20. Gerber M, Lang C, Lemola S, Colledge F, Kalak N, Holsboer-Trachsler E, Pühse U, & Brand S (2016). Validation of the German version of the insomnia severity index in adolescents, young adults and adult workers: Results from three cross-sectional studies. BMC Psychiatry, 16(1), 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, McLeod L, Delacqua G, Delacqua F, & Kirby J (2019). The REDCap consortium: Building an international community of software platform partners. Journal of Biomedical Informatics, 95, 103208. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, & Conde JG (2009). A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics, 42(2), 377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hu L. t., & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. [Google Scholar]
  24. Johns MW (1991). A new method for measuring daytime sleepiness: The Epworth Sleepiness Scale. Sleep, 14(6), 540–545. 10.1093/sleep/14.6.540 [DOI] [PubMed] [Google Scholar]
  25. Johnson DC, Polusny MA, Erbes CR, King D, King L, Litz BT, Schnurr PP, Friedman M, Pietrzak RH, & Southwick SM (2011). Development and initial validation of the Response to Stressful Experiences Scale. Military Medicine, 176(2), 161–169. [DOI] [PubMed] [Google Scholar]
  26. Kaufmann CN, Orff HJ, Moore RC, Delano-Wood L, Depp CA, & Schiehser DM (2019). Psychometric characteristics of the insomnia severity index in veterans with history of traumatic brain injury. Behavioral Sleep Medicine, 17(1), 12–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Koo TK, & Li MY (2016). A guideline of selecting and reporting intraclass correlation coefficients for reliability research. Journal of Chiropractic Medicine, 15(2), 155–163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lichstein KL (2017). Insomnia identity. Behaviour Research and Therapy, 97, 230–241. [DOI] [PubMed] [Google Scholar]
  29. Lin CY, Cheng AS, Nejati B, Imani V, Ulander M, Browall M, Griffiths MD, Broström A, & Pakpour AH (2020). A thorough psychometric comparison between Athens Insomnia Scale and Insomnia Severity Index among patients with advanced cancer. Journal of Sleep Research, 29(1), e12891. [DOI] [PubMed] [Google Scholar]
  30. Mindrila D. (2010). Maximum likelihood (ML) and diagonally weighted least squares (DWLS) estimation procedures: A comparison of estimation bias with ordinal and multivariate non-normal data. International Journal of Digital Society, 1(1), 60–66. [Google Scholar]
  31. Morin C. Insomnia Severity Index. Retrieved February from https://eprovide.mapi-trust.org/instruments/insomnia-severity-index#languages [Google Scholar]
  32. Morin CM (1993). Insomnia: Psychological Assessment and Management. The Guilford Press. [Google Scholar]
  33. Morin CM, Belleville G, Bélanger L, & Ivers H (2011). The Insomnia Severity Index: psychometric indicators to detect insomnia cases and evaluate treatment response. Sleep, 34(5), 601–608. 10.1093/sleep/34.5.601 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Morin CM, Colecchi C, Stone J, Sood R, & Brink D (1999). Behavioral and pharmacological therapies for late-life insomnia: a randomized controlled trial. JAMA, 281(11), 991–999. [DOI] [PubMed] [Google Scholar]
  35. Morin CM, Vallières A, & Ivers H (2007). Dysfunctional beliefs and attitudes about sleep (DBAS): Validation of a brief version (DBAS-16). Sleep, 30(11), 1547–1554. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Moscou-Jackson G, Allen J, Smith MT, & Haywood C Jr (2016). Psychometric validation of the insomnia severity index in adults with sickle cell disease. Journal of Health Care for the Poor and Underserved, 27(1), 209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Muraki E. (1992). A generalized partial credit model: Application of an EM algorithm. Applied Psychological Measurement, 16(2), 159–176. [Google Scholar]
  38. R Core Team. (2020). R: A language and environment for statistical computing. In https://www.R-project.org/
  39. Roane BM, & Taylor DJ (2008). Adolescent insomnia as a risk factor for early adult depression and substance abuse. Sleep, 31(10), 1351–1356. [PMC free article] [PubMed] [Google Scholar]
  40. Rogers AE, Caruso CC, & Aldrich MS (1993). Reliability of sleep diaries for assessment of sleep/wake patterns. Nursing Research, 42(6), 368–371. [PubMed] [Google Scholar]
  41. Rush AJ, Guillon CM, Basco MR, Jarrett RB, & Trivedi MH (1996). The Inventory of Depressive Symptomology (IDS): Psychometric properties. Psychological Medicine, 26, 477–486. 10.1017/S0033291700Q35558 [DOI] [PubMed] [Google Scholar]
  42. Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN, Markowitz JC, Ninan PT, Kornstein S, Manber R, Thase ME, Kocsis JH, & Keller MB (2003). The 16-item Quick Inventory of Depressive Symptomatology (QIDS) Clinician Rating (QIDS-C) and Self-Report (QIDS-SR): A psychometric evaluation in patients with chronic major depression. Biological Psychiatry, 54(5), 573–583. 10.1016/S0006-3223t02101866-8 [DOI] [PubMed] [Google Scholar]
  43. Samejima F. (1969). Estimation of latent ability using a response pattern of graded scores. Psychometrika monograph supplement. [Google Scholar]
  44. Samejima F. (1997). Graded response model. In Handbook of modern item response theory (pp. 85–100). Springer. [Google Scholar]
  45. Saunders JB, Aasland OG, Babor TF, De la Fuente JR, & Grant M (1993). Development of the alcohol use disorders identification test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption-II. Addiction, 88(6), 791–804. [DOI] [PubMed] [Google Scholar]
  46. Schlarb AA, Friedrich A, & Claßen M (2017). Sleep problems in university students–An intervention. Neuropsychiatric disease and Treatment, 1989-2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Smets EM, Garssen B, Bonke B, & De Haes JC (1995). The Multidimensional Fatigue Inventory (MFI) psychometric qualities of an instrument to assess fatigue. Journal of Psychosomatic Research, 39(3), 315–325. 10.1016/0022-3999(94)00125-0 [DOI] [PubMed] [Google Scholar]
  48. Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, & Jacobs GA (1983). Manual for the State-Trait Anxiety Inventory. Consulting Psychologists Press. [Google Scholar]
  49. Spielberger CD, Gorsuch RL, & Lushene RE (1970). The State-Trait Anxiety Inventory (Test Manual). Consulting Psychologists Press. [Google Scholar]
  50. Stephens RS, Roffman RA, & Curtin L (2000). Comparison of extended versus brief treatments for marijuana use. Journal of Consulting and Clinical Psychology, 68(5), 898. [PubMed] [Google Scholar]
  51. Taylor D, Wilkerson A, Pruiksma K, Dietch J, & Wardle-Pinkston S (2019). Structured Clinical Interview for Sleep Disorders-Revised (SCISD-R; ) https://insomnia.arizona.edu/SCISD [Google Scholar]
  52. Taylor DJ, Bramoweth AD, Grieser EA, Tatum JI, & Roane BM (2013). Epidemiology of insomnia in college students: relationship with mental health, quality of life, and substance use difficulties. Behavior Therapy, 44(3), 339–348. [DOI] [PubMed] [Google Scholar]
  53. van der Linden WJ, & Hambleton RK (2013). Handbook of modern item response theory. Springer Science & Business Media. [Google Scholar]
  54. Vector Psychometric Group, L. (n.d.). IRTPRO. In (Version 6.0) [Google Scholar]
  55. Veqar Z, & Hussain ME (2017). Validity and reliability of insomnia severity index and its correlation with pittsburgh sleep quality index in poor sleepers among Indian university students. International Journal of Adolescent Medicine and Health, 32(1). [DOI] [PubMed] [Google Scholar]
  56. Wilkerson A, Boals A, & Taylor DJ (2012). Sharpening our understanding of the consequences of insomnia: The relationship between insomnia and everyday cognitive failures. Cognitive Therapy and Research, 36(2), 134–139. [Google Scholar]
  57. Wong ML, Lau KNT, Espie CA, Luik AI, Kyle SD, & Lau EYY (2017). Psychometric properties of the Sleep Condition Indicator and Insomnia Severity Index in the evaluation of insomnia disorder. Sleep Medicine, 33, 76–81. [DOI] [PubMed] [Google Scholar]
  58. Yang CM, Wu CH, Hsieh MH, Liu MH, & Lu FH (2003). Coping with sleep disturbances among young adults: a survey of first-year college students in Taiwan. Behavioral Medicine, 29(3), 133–138. [DOI] [PubMed] [Google Scholar]
  59. Youden WJ (1950). Index for rating diagnostic tests. Cancer, 3(1), 32–35. [DOI] [PubMed] [Google Scholar]
  60. Zweig MH, & Campbell G (1993). Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clinical Chemistry, 39(4), 561–577. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1971885-supplement-1.docx (21.1KB, docx)
2
NIHMS1971885-supplement-2.docx (21KB, docx)
3
NIHMS1971885-supplement-3.docx (250.2KB, docx)
4
NIHMS1971885-supplement-4.docx (250.2KB, docx)

RESOURCES