Psychometric properties of the Turkish reduced morningness and eveningness questionnaire

ABSTRACT The aim of this study was to examine the psychometric properties of the 5-item Turkish Reduced Morningness-Eveningness Questionnaire (rMEQ) for the first time. The study involved 875 university students in an adaptation and validation study. Participants completed the rMEQ, MEQ, Depression Anxiety Stress Scale-21 (DASS-21), Insomnia Severity Index (ISI), Barratt Impulsiveness Scale Short Form (BIS-SF), and Oxford Happiness Questionnaire Short Form (OHQ-SF). The factor structure, convergent validity, internal consistency, sensitivity, and specificity of the rMEQ were examined. The confirmatory factor analysis showed that the rMEQ had a one-dimensional structure with good fit indices (χ2/df = 2.94, CFI = 0.990, TLI = 0.979, RMSEA = 0.047, and SRMR = 0.019). There was a significantly strong correlation between rMEQ and MEQ. In addition, we found a significantly weak correlation between rMEQ and DASS-21, ISI, BIS-SF, and OHQ-SF. The internal consistency coefficients of rMEQ were Cronbach’s α = 0.706 and McDonald’s ω = 0.740. The sensitivity and specificity of rMEQ were 83.3%–92.7% for morning types and 86.3%–87.3% for evening types. The Turkish rMEQ has adequate psychometric properties and can be used to assess an individual’s chronotype.


Introduction
The circadian rhythm is a 24-hour internal oscillation regulated by the suprachiasmatic nucleus in the hypothalamus (Danielsson et al. 2019).It controls sleep patterns, body temperature, hormonal secretions, circulation, and cognitive and physical performance.Circadian typology, or chronotype, refers to the best times for individuals to perform certain activities based on their natural sleep-wake preferences (Montaruli et al. 2021;Roenneberg et al. 2007).Chronotypes can vary greatly among people, creating a distribution of chronotypes ranging from larks to extreme owls, with the most people falling between these types (Roenneberg et al. 2016).More technically, chronotypes can be categorized into three groups: Morningness Type (MT), Eveningness Type (ET), and Intermediate Type (IT) (Adan et al. 2012).Approximately 60% of the adult population have the IT chronotype, while the remaining 40% have either the MT or ET chronotype (Adan et al. 2012).Individuals with ET tend to go to bed and wake up later than those with MT.Researchers have found that ET individuals are at risk for mental health problems such as depression, suicide attempts, sleep disturbances, impulsivity, eating disorders, and personality disturbances, as well as physical diseases such as hypertension, obesity, cardiovascular diseases, fibromyalgia, and diabetes.In contrast, MT individuals tend to have healthy habits and a healthy body (Bauducco et al. 2020;Hashemipour et al. 2020;Kantermann et al. 2012;Kivelä et al. 2018;Li et al. 2020;Makarem et al. 2020;Roenneberg et al. 2016;Schulte and Riemersma-van der Lek 2021;Taylor and Hasler 2018;Urbán et al. 2011).Roenneberg found a more significant difference between the biological and social clock in people with ET than in people with MT.This difference is known as social jet lag.Social jet lag is associated with obesity, lower academic performance, and higher levels of depression in university students (Roenneberg et al. 2019).ET exhibited lower happiness levels than MT in a research study of 564 pre-clinical medical students in Turkey (Tan et al. 2020).In a different research, including 1035 university students in Turkey, individuals with ET had higher levels of food addiction, psychological pain, and social jetlag (Ceylan et al. forthcoming).An additional study comprising 902 university students in Turkey showed a link between problematic smartphone use and being an ET (Demirhan et al. 2016).
Several surveys have been developed to assess an individual's chronotype.The most common ones are the Morningness-Eveningness Questionnaire (MEQ), reduced Morningness-Eveningness Questionnaire (rMEQ), Munich Chronotype Questionnaire (MCTQ), Morningness-Eveningness Stability Scale Improved (MESSi), and Composite Scale of Morningness (CSM).The MEQ is a widely used and recognized tool for this purpose.To date, MEQ's original study has been cited 7260 times (Caci et al. 2009;Kerkhof 1985;Tankova et al. 1994).The MEQ was developed by Horne and Östberg in 1976 and is considered the gold standard for assessing chronotype (Horne and Ostberg 1976).It has been translated into various languages and has shown strong reliability and validity in numerous studies across different countries including Slovenia, Italy, Turkey, France, India, Sweden, Korea, Italy (Agargun et al. 2007;Di Milia et al. 2013;Horne and Ostberg 1976;Lee et al. 2014;Roveda et al. 2017;Taillard et al. 2004;Treven Pišljar et al. 2019;Zacharia et al. 2014).The MEQ consists of 19 items that evaluate sleep-wake patterns, optimal times for physical and mental activities, and subjective alertness using a Likert-type response format (Horne and Ostberg 1976).Studies have proposed MEQ cut-off scores to classify individuals as MT, IT, or ET.
In Turkey, there are four instruments for determining adult chronotypes: MEQ, MCTQ, MESSi and CSM.The MEQ is the oldest and most widely used instruments for chronobiological research in Turkey (Balcioglu et al. 2022;Beşoluk et al. 2011;Kandeger et al. 2018Kandeger et al. , 2019;;Kurnaz and Kandeger 2020;Özçelik et al. 2023;Selvi et al. 2007Selvi et al. , 2017)).MEQ consists of 19 items, MCTQ of 31 items, MESSi of 15 items, and CSM of 13 items.Some researchers have developed singleitem scales instead of multi-item scales.They reported that participants perceived multi-item scales as repetitive and onerous, resulting in low response rates.On the other hand, it is recommended that the number of total items on the scale be at least 3 or 4 (Robinson 2017).More items will be needed to capture the richness of multidimensional variables (Allen and Meyer 1990).However, this must be balanced with the need for scale brevity to maximize response rates.For this reason, short versions of well-known scales are often developed (Thompson 2007).In chronobiology instruments, only short versions of the MEQ and MCTQ have been created (Adan and Almirall 1991;Ghotbi et al. 2020).The psychometric properties of the Turkish version of neither of these short versions have been examined.Shortening scales can be advantageous in research involving larger populations, as shorter scales are more likely to be completed by participants, leading to better compliance (Di Milia et al. 2013;Kemper et al. 2019).The field of chronobiology requires shorter scales due to the complexity of longer ones, which pose challenges for data modeling (Ziegler et al. 2014).Adan and Almirall developed a shortened version of the MEQ, known as the rMEQ, which includes only items 1, 7, 10, 18, and 19 from the original MEQ (Adan and Almirall 1991).The rMEQ has gained popularity due to its brief nature and strong psychometric qualities.It has been assessed for psychometric properties in several countries, including Hungary, Sweden, France, Spain, America, Italy, the Kingdom of Saudi Arabia, China, Germany, Poland, India, Iran, and Portugal (Adan and Almirall 1991;BaHammam et al. 2011;Biswas et al. 2014;Caci et al. 2009;Carciofo et al. 2012;Chelminski et al. 2000;Danielsson et al. 2019;Jankowski 2013;Loureiro and Garcia-Marques 2015;Natale et al. 2006;Rahafar et al. 2014;Randler 2013;Urbán et al. 2011).Studies on the rMEQ have consistently shown a unidimensional factor structure, good internal consistency, robust convergent and construct validity, and good test-retest reliability (Biswas et al. 2014;Danielsson et al. 2019;Natale et al. 2006;Rahafar et al. 2014).
To the best of our knowledge, the psychometric properties of the Turkish version of the rMEQ have not been evaluated.Therefore, this study aims to assess the psychometric properties of the Turkish rMEQ version for the first time.Our goal is to determine the factor structure, convergent validity, reliability, sensitivity, and specificity of the Turkish rMEQ.We also looked at the relationship between chronotypes and levels of depression, anxiety, happiness, impulsivity, and insomnia, as these are the most commonly investigated relationships in chronotype studies.

Study design
This study was an adaptation and validation study conducted at Atatürk University in Turkey.University students participated in a paper-based survey between January 15, 2023, and March 15, 2023, after providing informed consent.We randomly recruited university students who could read and write Turkish and allowed them to freely participate in the response process.Participants were excluded if they were on leave, out of school, or chose not to engage in the study.For the study, we developed a psychological battery, which included the sociodemographic sleep form, the Morningness-Eveningness Questionnaire (MEQ), the Reduced Morningness-Eveningness Questionnaire (rMEQ), the Insomnia Severity Index (ISI), the Oxford Happiness Questionnaire Short Form (OHQ-SF), the Depression Anxiety Stress Scale-21 (DASS-21), and the Barratt Impulsiveness Scale Short Form (BIS-SF).DASS-21, ISI, BIS-SF, and OHQ-SF are the most commonly used instruments in the literature to determine levels of depression, anxiety, stress, insomnia, impulsivity, and happiness, respectively.Therefore, we used these scales in addition to rMEQ and MEQ.The instruments were administered during class breaks to allow participants to complete the questionnaires.The instruments' order was as follows: sociodemographic sleep form, MEQ, ISI, OHQ-SF, DASS-21, BIS-SF, and rMEQ.The research team waited in the classroom while the participants filled in the scales.If the participants had any questions about the research, our research team answered them there.The study was conducted in accordance with the Declaration of Helsinki and was approved by the Atatürk University Faculty of Medicine Non-Invasive Clinical Ethics Committee (approval date: 29.12.2022, number: 40).

Adaptation of Turkish rMEQ
The adaptation and validation study of the Turkish 19item MEQ was conducted (Agargun et al. 2007).Since the rMEQ consists of 5 items of MEQ, we did not carry out an adaptation process in our study as in a classical adaptation study.Our research team checked all items in the rMEQ for comprehensibility, and we did not find any problems with them.While the scale was administered to the participants, there were no objections regarding the scale items.We developed the Turkish rMEQ using the Turkish MEQ as a baseline, comprising only items 1, 7, 10, 18, and 19 from the original long MEQ version.Therefore, we selected the first, seventh, tenth, eighteenth, and nineteenth items from the Turkish MEQ and included them in the Turkish rMEQ version.

Sample size
The rMEQ has five questions, and for the validityreliability study of the rMEQ, it is recommended to have a minimum of 15-100 participants based on the number of participants per item.While it is typically advised to have 3-20 participants per item in validity and reliability assessments of a scale, there has been a recent desire to increase the number of participants to near 1000 (Boateng et al. 2018;Comrey 1988;Gunawan et al. 2021;White 2022).Our study included 875 participants, meeting the recommended sample size for the psychometric analysis of rMEQ.

Sociodemographic sleep form
The sociodemographic sleep form consists of two parts.In the first part, participants completed a self-reported questionnaire that included demographic questions about age, gender, weight, height, residential status, nicotine and alcohol use, and chronic diseases.In the second part, participants answered four questions about their sleep patterns: (  (Horne and Ostberg 1976).Turkish MEQ conducted adaptation and validation; the test-retest correlation coefficient is 0.84 (Agargun et al. 2007).Our study found a robust Cronbach's α of 0.853 for the Turkish MEQ.

The rMEQ
The rMEQ is a questionnaire with five items, which are items 1, 7, 10, 18, and 19 of the MEQ (Adan and Almirall 1991).We obtained the Turkish rMEQ version by using items 1, 7, 10, 18, and 19 from the Turkish MEQ.The rMEQ total scores range from 4-25.Individuals with scores below 12 are classified as ET, scores between 12 and 17 are classified as IT, and scores over 17 are classified as MT (Danielsson et al. 2019).The first item of rMEQ is the preferred time to wake up on free days in the morning.The second item is: How tired do you feel in the first half hour after waking up in the morning?The third item is: At what time in the evening do you feel tired and need sleep?The fourth item is: At what time of the day do you think that you reach your "feeling best" peak?And the fifth item: One hears about "morning" and "evening" types of people.Which one of these types do you consider yourself to be?In the first, third, and fourth items, time periods are included in 5-point Likert responses, and each question is scored between 1 and 5 points.The second and fifth items consist of a 4-point Likert scale.In the second item, responses between very tired and very refreshed are scored between 1 and 4 points.The fifth item assesses which chronotype the respondent feels close to.There are four responses between definitely a "morning" type (6 points) and definitely a "evening" type (0 points).The original English and Turkish versions of the rMEQ are presented in Appendix.

The Insomnia Severity Index (ISI)
The ISI is a widely used tool for measuring sleep disturbance (Bastien et al. 2001;Boysan et al. 2010).It consists of seven items and uses a 5-point likert scale to evaluate specific areas of insomnia, including the severity of initial, middle, and late insomnia, contentment with sleep, impact on daytime functioning, noticeability of sleep problems by others, and distress caused by sleep difficulties.The total score ranges from 0 to 28, with higher scores indicating a higher level of insomnia.
The Turkish ISI demonstrates a two-factor structure and has shown adequate internal consistency with a coefficient of 0.79 (Boysan et al. 2010).In our study, the Turkish version of the ISI showed a satisfactory Cronbach's α of 0.797.

The Oxford Happiness Questionnaire Short Form (OHQ-SF)
The original OHQ-SF version consists of eight items and uses a 6-point Likert-type scale (Hills and Argyle 2002).The Turkish OHQ-SF, on the other hand, is a 7-item measure designed to assess happiness levels (Doğan and Çötok 2011).It utilizes a 5-point Likerttype scale, with response options ranging from 1 (Never Disagree) to 5 (Totally Agree), making it short and easy to administer.The Turkish OHQ-SF demonstrates a single-factor structure and has shown adequate internal consistency with a coefficient of 0.74 (Doğan and Çötok 2011).The total score on the OHQ-SF ranges from 7 to 35, with higher scores indicating a higher level of happiness.In our study, the Turkish OHQ-SF exhibited a robust Cronbach's α of 0.789.

The Depression Anxiety Stress Scale-21 (DASS-21)
The DASS-21 is a widely used tool to measure depression, anxiety, and stress (Lovibond and Lovibond 1995;Oei et al. 2013).It consists of 21 items, each scored from 0 to 3, resulting in a total score ranging from 0 to 21.
Higher scores indicate higher levels of depression, anxiety, and stress (SariçSariçAm 2018; Lovibond and Lovibond 1995).The Turkish DASS-21 cronbach's alpha internal consistency reliability coefficient was found to be α = 0.87 for the depression subscale, α = 0.85 for the anxiety subscale, and α = 0.81 for the stress subscale (SariçSariçAm 2018).The DASS-21 has been frequently used in research to assess the severity of these mental health dimensions (Adhikari et al. 2023;Oei et al. 2013;Uygur et al. 2022).In our study, the Turkish version of the DASS-21 showed excellent internal consistency with a Cronbach's α of 0.91.

The Barratt Impulsiveness Scale Short Form (BIS-SF)
The BIS-SF is a 15-item self-report questionnaire that measures impulsivity.It uses a 4-point Likert scale (rarely/never, sometimes, often, almost always/always).
Higher BIS-SF scores indicate higher levels of impulsivity.Tamam et al. (2013) conducted a validity and reliability analysis of the Turkish BIS-SF and found a Cronbach's alpha value of 0.82, indicating strong psychometric properties (Tamam et al. 2013).In our study, we also found a robust Cronbach's α (0.827) for the Turkish BIS-SF.

Statistical analysis
The data were analyzed using SPSS (version 20.0), SPSS (Amos 24), and Jamovi (version 2.3.28).We used SPSS (version 20.0) for descriptive statistics and correlation analyses, SPSS (Amos 24) for factor structure, and Jamovi (version 2.3.28) for reliability analyses, medical decision analyses, sensitivity, specificity, positive predictive value, and negative predictive value.The descriptive statistics displayed the mean, standard deviation, and frequency.Skewness and kurtosis scores were used to assess the normality of the data.Confirmatory factor analysis (CFA) was conducted to evaluate the factor structure using the maximum likelihood estimation approach.P-values below 0.05 were considered statistically significant.We used a CFA to assess the structural validity of the rMEQ, aiming to determine how accurately the scores represented the underlying dimensions.Since the rMEQ is usually thought of as a single-factor scale, we used a single-factor model to check its structural validity (Biswas et al. 2014;Chelminski et al. 2000;Danielsson et al. 2019;Natale et al. 2006;Rahafar et al. 2014).The model's fit was evaluated based on the following criteria: a chi-square/degree of freedom (χ2/df) ratio of less than 3.0, a root-mean-square-error of approximation (RMSEA) of less than 0.06, a comparative fit index (CFI) of greater than 0.90, a Tucker-Lewis index (TLI) of greater than 0.95, and a standardized root-mean-square residual (SRMR) close to or below 0.07 (Bentler 1980(Bentler , 1990;;Brown 2015;Hu and Bentler 1999;Schermelleh-Engel et al. 2003;Sun 2005).To assess the convergent validity of the rMEQ, we calculated Pearson correlation coefficients for the rMEQ with MEQ, ISI, DASS-21, BIS-SF, and OHQ-SF (Humphreys et al. 2019).We also examined inter-item and item-total correlations.We hypothesized a strong correlation (r = 0.700-0.900)between the rMEQ and the MEQ.Additionally, we expected higher scores on the rMEQ (indicating morningness preference) to be significantly negatively correlated with ISI, DASS-21, and BIS-SF and significantly positively correlated with OHQ-SF.
We assessed the internal consistency of the rMEQ by calculating Cronbach's α and McDonald's ω coefficients.A coefficient of 0.700 or higher was considered acceptable (Cronbach 1951;Taylor 2021).We considered MEQ as the gold standard test for determining MT, ET, and IT.We compared the overlap between rMEQ and MEQ for participants with different chronotypes to assess how well rMEQ identified MT, ET, and IT.Additionally, we calculated the sensitivity, specificity, positive predictive value, and negative predictive value of rMEQ to determine its diagnostic power in detecting chronotypes.We categorized participants into MT and ET groups based on their rMEQ scores.We used independent t-tests and Chi-square tests to compare the total scores of all scales and sociodemographic sleep variables between the MT and ET groups.

Reliability analysis of the rMEQ
The mean scores for the items on the rMEQ scale were as follows: 2.76 ± 1.107, 2.38 ± 0.684, 3.00 ± 1.102, 2.79 ± 0.959, and 3.15 ± 1.862.The internal consistency of the rMEQ was found to be Cronbach's α = 0.706 and McDonald's ω = 0.740, which were considered sufficient (Cronbach 1951;Taylor 2021).Cronbach's α ve McDonald's ω analyzed from the total rMEQ increased slightly after removing item 3 (the value of the alpha and omega if items dropped ranged between 0.60-0.71and 0.63-0.75)(see Table 5).The inter-item correlations of the rMEQ ranged from 0.09 to 0.57, and the item-total correlations ranged from 0.52 to 0.86 (see Figure 1).chronotypes.Using MEQ as the gold standard test, rMEQ correctly identified 797 out of 875 participants as MT, resulting in a sensitivity of 83.3% and a specificity of 92.7%.For ET, rMEQ correctly identified 762 out of 875 participants, with a sensitivity of 86.3% and a specificity of 87.3%.Regarding IT, rMEQ correctly identified 684 out of 875 participants, with a sensitivity of 73.8% and a specificity of 85%.The Youden index of rMEQ was 0.76, 0.73, and 0.58 for MT, ET, and IT, respectively, indicating that the diagnostic power of rMEQ was highest for MT, followed by ET and IT.

Comparison of variables between MT and ET groups
Table 8 presents a comparison of variables between the MT and ET groups.There were no differences in age, gender, and BMI between the two groups.However, smoking and alcohol use were significantly higher in the ET group compared to the MT group.Additionally, the ET group reported more insomnia symptoms and greater use of medication for insomnia in the last month.The ET group also had less sleep time and reported significantly worse sleep quality.On the other hand, the ET group had significantly higher mean scores for DASS-21, ISI, and BIS-SF, while the MT group had significantly higher scores on the OHQ-SF.

Discussion
Circadian preferences, known as chronotypes, are closely linked to various health conditions and sleep-wake habits (Rhee et al. 2012;Roenneberg et al. 2007;Taylor and Hasler 2018).Therefore, assessing an individual's chronotype is essential for understanding circadian rhythms and associated disorders.The Morningness-Eveningness Questionnaire (MEQ) is the most commonly used tool for evaluating circadian preferences (Horne and Ostberg 1976).The short form of the MEQ (rMEQ) is more practical for participants and suitable for studies with larger samples.However, the psychometric properties of the rMEQ have not been examined in Turkey.The main aim of our study was to evaluate the psychometric properties of the Turkish rMEQ.We examined the factor structure, convergent validity, reliability, sensitivity, and specificity of the rMEQ.Additionally, we investigated whether there is a positive relationship between the eveningness type and severe insomnia, excessive impulsivity, high depression, anxiety, stress symptoms, and low levels of happiness in the Turkish rMEQ.Our study's first main finding is that the rMEQ exhibited a single-factor structure, consistent with previous studies (Adan and Almirall 1991;Caci et al. 2009;Chelminski et al. 2000;Jankowski 2012;Rahafar et al. 2014;Randler 2013).Previous studies have used exploratory factor analysis (EFA) to examine the factor structure of the rMEQ (Chelminski et al. 2000;Danielsson et al. 2019;Jankowski 2013;Randler 2013), with only one study discussing the factor structure of the rMEQ using CFA.This study found a two-factor structure for the rMEQ with CFA (Urbán et al. 2011), while our CFA analysis showed that the rMEQ exhibited a one-factor structure with good fit indices.CFA analysis is recommended if there is a strong theory about the scale's factor structure (Kline 2011;Orcan 2018).The consistent single-factor structure observed in most EFA studies led us to believe that the factor structure of the rMEQ should be examined using CFA (Adan and Almirall 1991;Caci et al. 2009;Danielsson et al. 2019;Jankowski 2013;Rahafar et al. 2014;Randler 2013).In the most recent validation study of the rMEQ, Danielsson et al. (2019) also recommended examining the factor structure of the rMEQ with CFA.
Our study's second main finding is the significant correlations detected in the convergent validity analysis.Previous studies have found strong correlations between rMEQ and MEQ (0.89, 0.90, 0.88) (Caci et al. 2009;Carciofo et al. 2012;Jankowski 2013).Consistent with these results, we also found a strong correlation (0.87) between rMEQ and MEQ, indicating that Turkish rMEQ can be used instead of Turkish MEQ.It is welldocumented in the literature that ET is associated with higher levels of depression, anxiety, stress, impulsivity, and insomnia severity (Antypa et al. 2016;Hidalgo et al. 2009;Passos et al. 2017;Randler 2011;Taylor and Hasler 2018).Our study found significant negative correlations between rMEQ (lower scores indicating ET) and DASS-21, BIS-SF, and ISI, suggesting that individuals with ET have higher levels of these adverse mental health symptoms.Additionally, we found a significant positive correlation between rMEQ and OHQ-SF,  indicating that individuals with MT have higher levels of happiness.These findings are consistent with previous research (Danielsson et al. 2019;Rahafar et al. 2014).Unlike many rMEQ validation studies (Carciofo et al. 2012;Jankowski 2013;Randler 2013) that only examined the correlation between rMEQ and MEQ, our study used different scales to measure variables such as depression, anxiety, stress, insomnia severity, impulsivity, and happiness levels.The correlations we found between rMEQ scores and other scale scores demonstrate that ET is associated with adverse mental health symptoms and MT with positive mental health symptoms, supporting the use of Turkish rMEQ in chronotype studies.
The third main finding of our study is that the rMEQ had an internal consistency coefficient of 0.70 for Cronbach's α and 0.74 for McDonald's ω, indicating adequate internal consistency (Cronbach 1951;Taylor 2021).Previous studies only examined Cronbach's α value for the internal consistency of the rMEQ (Caci et al. 2009;Carciofo et al. 2012;Danielsson et al. 2019;Jankowski 2013;Rahafar et al. 2014;Randler 2013).In contrast, we also looked at the McDonald's ω value and found that the scale showed adequate internal consistency in this value.The Cronbach's alpha values of the rMEQ in other studies ranged between 0.68 and 0.78 (Caci et al. 2009;Carciofo et al. 2012;Danielsson et al. 2019;Jankowski 2013;Rahafar et al. 2014;Randler 2013).In a study by Urbán et al. (2011), Cronbach's α of the rMEQ was found to be relatively low at 0.59.However, the mean age of the sample in that study was relatively low (15.3 years) compared to 21.42 years in our study.Some additional questions were added to the rMEQ in that study (Urbán et al. 2011).These reasons may have led Urbán et al. (2011) to find a low internal consistency coefficient.We found a weak interitem correlation between item 2 and item 3, which was also found in other studies (Danielsson et al. 2019;Loureiro and Garcia-Marques 2015;Randler 2013).However, in our study, the correlations between itemtotal items were not low.When we looked at the changes in the internal consistency coefficients when item 3 was removed, we found that Cronbach's α and McDonald's ω increased slightly.However, since the factor loading of item 3 was greater than 0.30 and the correlation between item 3 and the total item was 0.55, we did not consider removing this item.A similar result was found in the Persian version of the rMEQ.In that study, Cronbach's α increased slightly when item 3 was removed.However, in that study, item 3 was not removed because the factor loading of item 3 was sufficient (Rahafar et al. 2014).In addition, we found that items 2 and 3 had lower factor loadings, but these values were above 0.30.This may be related to cultural and language differences.
The fourth main finding of our study was to determine the diagnostic power of the rMEQ.Previous studies have used actigraphy, MEQ, or the Composite Scale of Morningness to assess the power of rMEQ to identify chronotypes (Caci et al. 2009;Carciofo et al. 2012;Jankowski 2013;Randler 2013;Tonetti and Natale 2019).Our study compared the rMEQ with the MEQ, the most widely used scale for chronotype identification.According to the MEQ, 21.7% of all participants were ET, 61.1% IT, and 17.1% MT.However, according to rMEQ, 28.7% were ET, 50.9% IT, and 20.3% MT.The distribution of participants according to both scales was IT>ET>MT, which is consistent with other studies with healthy participants (BaHammam et al. 2011;Chelminski et al. 2000;Natale et al. 2006).However, in some studies, the distribution was IT>MT>ET (Loureiro and Garcia-Marques 2015;Urbán et al. 2011).The studies that found this result were the Hungarian and Portuguese studies (Loureiro and Garcia-Marques 2015;Urbán et al. 2011).Factors such as sunrise time, culture, and climate can affect chronotypes.People in warmer climates tend to be more MT.People also tend to be more MT where the sunrise time is earlier (Danielsson et al. 2019).Also, where the culture is different, for example, in Saudi Arabia, where it is critical to pray in the morning, MT is higher (BaHammam et al. 2011).Our study was conducted in Erzurum, located in northeastern Turkey, which has a frigid climate that may have influenced the distribution of chronotypes.The cold climate may have caused more individuals with ET than individuals with MT in our study.However, the fact that the sun rises early and morning prayers are important in Erzurum, as in Saudi Arabia, suggests that more complex processes may impact chronotypes.When we compared the rMEQ and MEQ, there was no transitivity between the ET and MT groups on either scale.The chronotype group of 78.8% of the participants did not change according to rMEQ in our study, which is consistent with previous findings.This rate was very close to the rates found by Adan and Almirall (78% unchanged), Chelminski et al. (80% unchanged), and Carciofo (83.3% unchanged) (Adan and Almirall 1991;Carciofo et al. 2012;Chelminski et al. 2000).The sensitivity of rMEQ was ET > MT > IT, and its specificity was MT > ET > IT.This study shows that rMEQ successfully detects individuals with ET and MT chronotypes.
Finally, our study used the rMEQ to determine the chronotypes of the participants and compared the MT and ET groups based on sociodemographics, sleep variables, and scales.e found no significant differences in age, gender, or BMI between the MT and ET groups.Previous studies have suggested that ET is more common in young men, possibly due to factors such as gaming, problematic social media use, unhealthy sleep habits, smoking, and alcohol consumption (Ceylan et al. forthcoming;Danielsson et al. 2019;Ghotbi et al. 2023;Randler 2013;Randler et al. 2016;Urbán et al. 2011;Üzer et al. forthcoming).However, our study and others did not find a difference in chronotype between males and females (Caci et al. 2009;Jankowski 2013;Loureiro and Garcia-Marques 2015).ET has been associated with lower dietary restriction, less healthy eating habits, and a higher BMI (Lucassen et al. 2013;Roenneberg 2023).In our study, the BMI of the ET group was higher but not significantly different from the MT group, possibly due to similar eating habits among university students.The rate of smokers and alcohol users was higher in the ET group, consistent with previous studies (Adan 1994;Urbán et al. 2011).Those with ET in our study also exhibited higher levels of depression, anxiety, stress, insomnia severity, impulsivity, and lower levels of happiness, consistent with previous research (Bauducco et al. 2020;Lee et al. 2014;Li et al. 2020;Lucassen et al. 2013;Roenneberg 2023;Selvi et al. 2011;Tan et al. 2020;Taylor and Hasler 2018).In our study, sleep quality was lower, and sleep duration was shorter in the ET group, with higher rates of insomnia symptoms and medication use for insomnia.ET may predispose to depression and less happiness with shorter sleep duration and poorer sleep quality (Kalmbach et al. 2018).ET has not been suggested as a risk factor for anxiety disorders, with mixed findings in previous studies (Antypa et al. 2016).This may be due to the different study samples (Danielsson et al. 2019;Passos et al. 2017;Selvi et al. 2012).High impulsivity levels are positively correlated with ET, possibly due to sleep deprivation or cognitive deficiencies in ET patients (Selvi et al. 2011;Wang et al. 2022).These findings show that disruption of the circadian rhythm is associated with some health problems.Regulating the circadian rhythm may be an essential step in preventing these health problems (Roenneberg et al. 2022).To regulate the circadian rhythm, especially sleep, it is necessary to determine the individual's chronotype and organize sleep hours accordingly.Implementing flexible working hours and later school start times for university students could regulate the circadian rhythm.

Limitations
This study has several limitations.Firstly, using selfreports to examine the psychometric properties of the Turkish rMEQ may have introduced response bias in the participants.Secondly, the study focused on young participants, who are more likely to experience sleep problems and delayed sleep-wake cycles, potentially impacting the results.Thirdly, a re-test analysis for the Turkish rMEQ was not conducted.Fourthly, we did not look at the relationship between chronotypes and the Big Five factors in the other chronotype studies.Lastly, applying multiple scales to participants may have reduced the response quality.Research on whether survey lengths affect response rates and response quality needs to be more consistent.Informing participants in advance, following up with those who do not respond, and making it enjoyable can increase the response rate and response quality (Sahlqvist et al. 2011).The application of many surveys in our study may have caused the number of participants to decrease.At the same time, it may have led to monotonous responses -that is, poorquality responses.However, we informed the participants before administering the survey, stood by them, and helped them while administering the survey.In this respect, our response quality is higher than that of an online survey.
Despite these limitations, our study is unique in that it is the first to examine the psychometric properties of the Turkish rMEQ, has a good sample size, utilizes CFA to explore the factor structure of the rMEQ, and includes other scales measuring various psychiatric symptoms.Future studies could validate the Turkish rMEQ using actigraphy, sleep diaries, body temperature, or melatonin level measures.Additionally, future validation studies of the rMEQ in Turkey should include different age groups and conduct re-test analysis.

Conclusion
The Turkish rMEQ demonstrated a one-factor structure and adequate internal consistency, so it can be considered to have satisfactory psychometric properties.Additionally, it showed high similarity to the original MEQ.As a result, the rMEQ can be used as a suitable alternative to the MEQ in chronobiology studies in Turkey.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The author(s) reported there is no funding associated with the work featured in this article.

Table 1 .
The sociodemographic and sleep characteristics of all participants (n = 875).

Table 2 .
Ages of participants and total scores of all scales (N = 875).

Table 3 .
Model fit indices and factor loadings of the single-factor model for the rMEQ (N = 875).

Table 4 .
Correlations between rMEQ and other scales.

Table 6 .
Numbers and percentages of chronotypes according to rMEQ and MEQ total scores.

Table 7 .
Overlap of chronotype groups between rMEQ and MEQ.

Table 8 .
Comparison of variables between MT and ET groups.