Single Item Stress Question

1

5 (Likert skala):
1= Inte alls; 2= bara lite; 3=I viss mån; 4=ganska mycket; 5=väldigt mycket.

Summera.

Personer som svarar 4 eller 5 under 2-4 på varandra följande veckor kan vara i risk för framtida depression, utmattning och sjukskrivning.

Frågan ställs en gång per vecka via SMS. Påminnelse skickas efter 48 timmar vid uteblivet svar.

Utskick av frågor, insamling av svar samt utskick av påminnelser sköts automatiskt av SMS systemet, dock tar uppstarten av systemet viss tid (lägga in ID nummer, telefon, anvisningar, mm).

Svensk version av frågan:
Med stress menas en situation då man känner sig spänd, rastlös, nervös eller orolig eller inte kan sova på natten eftersom man tänker på problem hela tiden.
Har du känt dig stressad den senaste tiden?

Version på engelska:
“Stress means a state in which a person feels tense, restless, nervous or anxious or is unable to sleep at night because his/her mind is troubled all the time. Do you feel this kind of stress these days?”

18-67

r = 0.804-0.868

Ej relevant eftersom det var EN fråga. Nedan följer dock tilläggs information kring reliabilitet (från vår studie, utöver det som står i referensartikeln):

To estimate the reliability of the SISQ with the correction for attenuation formula (Nunnally & Burnstein, 1994), we used the data collected at the two baseline measurements (four months apart); only employees who answered both the ERI questionnaire and the SISQ were selected (Baseline 1 N=91; Baseline 2 N=97).
A further way to estimate the reliability of the SISQ was principal axis factor analysis (PAF). The data came from the two baseline measurements but the N differed slightly from the above because not all employees answered the SISQ. For baseline 1, N=84, for baseline 2, N=93.

The estimation of the reliability of the SISQ according to the correction for attenuation formula followed procedures described in studies by Wanous and Reichers(Wanous & Hudy, 2001; Wanous & Reichers, 1996). We used the over-committment subscale for calculations because it demonstrated the highest correlation with the SISQ when we examined for convergent validity. The correction for attenuation formula can be used irrespective of whether variables/scales come from the same or from different conceptual domains (Wanous & Hudy, 2001). If the underlying true correlation between two variables is 1, the observed correlation between the same two variables is only limited by their reliabilities (ibid.). If an observed correlation and the reliability of one variable is known, it is possible to assess the reliability of the other variable by solving the equation The assumption that the SISQ and the over-commitment subscale of ERI are from the same domain was relaxed to the underlying true correlation of .90, and then to .80. The procedure was carried out both with the data from baseline 1 and baseline 2. The average of the two was then calculated.

PAF was computed for the over-commitment scale plus the SISQ. The communality of the SISQ with items in another scale from the stress domain (in this case the over-commitment subscale of the ERI) can be seen as a demonstration of its reliability, in accordance with the view of Wanous and Hudy (2001) that the communality can be seen as “a conservative estimate of single-item reliability”. If the SISQ loads on the same factor as the items of the over-commitment subscale, the communality of the SISQ can be seen as an estimate of its reliability. It could also strengthen the assumption used in the assessment of reliability with the correction for attenuation formula (i.e. that the SISQ and over-commitment come from same conceptual domain). The procedure was carried out with the data from baseline 1 and baseline 2, after which the average of the two was computed. Finally, we computed the average of the attenuation formula and the factor analysis.

Reliability Estimated Using The Correction for Attenuation Formula. The alpha reliability for the over-commitment scale was .86 for the baseline measurement 1, and .83 for the baseline measurement 2. The correlation between the over-commitment scale and the SISQ was .72 for the baseline 1 and .62 for the baseline 2. Accordingly, the minimum reliability for the SISQ is .74 for baseline 1 and .57 for baseline 2, if underlying correlation is assumed to be .90. If the true underlying correlation is assumed to be .80, the minimum reliability of the SISQ is .94 and .73 for the baseline 1 and 2 respectively. The average estimate of the SISQ reliability is .66 if computed on an underlying true correlation of .90, and .84 if computed on an underlying true correlation of .80.

Reliability estimated using PAF. The communalities of the SISQ and the items of the over-commitment subscale were estimated using PAF and are shown in Table 5. The SISQ has higher communality than three of the items in the over-commitment scale in baseline 1, and is near the scale average in baseline 2.

Vid valideringen användes SMS-Track Questionnaire software system 1.1.3 (2007) för att distribuera frågan och samla in svaren. För mer info se https://www.sms-track.com/
Andra likvärdiga system kan användas.

Medelåldern i valideringsdelen av studien var 44.6 år (11.7). Medianen för SISQ var 3. Övervägande delen av valideringspopulationen var universitetsutbildade kvinnor. Populationen: primärvårdsanställda.

Bozana Johansson
leg.psykolog, leg.psykoterapeut, doktorand
Karolinska Institutet
The Institute of Environmental Medicine (IMM)
Unit of Intervention and Implementation Research

Work-related stress assessed by a text message single-item stress question
==========================================================================

Background Given the prevalence of work stress-related ill-health in the Western world, it is important to find
cost-effective, easy-to-use and valid measures which can be used both in research and in practice.

Aims To examine the validity and reliability of the single-item stress question (SISQ), distributed weekly
by short message service (SMS) and used for measurement of work-related stress.

Methods The convergent validity was assessed through associations between the SISQ and subscales of the
Job Demand–Control–Support model, the Effort–Reward Imbalance model and scales measuring
depression, exhaustion and sleep. The predictive validity was assessed using SISQ data collected
through SMS. The reliability was analysed by the test–retest procedure.

Results Correlations between the SISQ and all the subscales except for job strain and esteem reward were
significant, ranging from −0.186 to 0.627. The SISQ could also predict sick leave, depression and
exhaustion at 12-month follow-up. The analysis on reliability revealed a satisfactory stability with a
weighted kappa between 0.804 and 0.868.

Conclusions The SISQ, administered through SMS, can be used for the screening of stress levels in a working
population.

Nunnally, J., & Burnstein, I. (1994). Psychometric Theory. McGraw-Hill: New York.
Wanous, J. P., & Hudy, M. J. (2001). Single-Item Reliability: A Replication and Extension. Organizational Research Methods, 4(4), 361-375. doi: 10.1177/109442810144003
Wanous, J. P., & Reichers, A. E. (1996). Estimating the reliability of a single-item measure. Psychological Reports, 78(2), 631-634.
Arapovic-Johansson, B., Wåhlin, C., Kwak, L., Björklund, C., & Jensen, I. (2017). Work-related stress assessed by a text message single-item stress question. Occupational Medicine. doi: 10.1093/occmed/kqx111
Elo, A.-L., Leppänen, A., & Jahkola, A. (2003). Validity of a single-item measure of stress symptoms. Scandinavian Journal of Work, Environment & Health, 29(6), 444-451. doi: 10.5271/sjweh.752

Bozana Johansson

http://ki.se/en/people/bonjoh