INTRODUCTION

Despite sustained interest from medical education researchers for over 50 years, evaluating medical school learning environments remains challenging. At least 15 tools have been used to assess undergraduate medical school learning environments, but none have strong evidence for validity [1]. Our team of researchers recently developed a new learning environment scale, called the Johns Hopkins Learning Environment Scale (JHLES), at the Johns Hopkins University School of Medicine (JHUSOM), which, after exploratory factor analysis on data from JHUSOM students, yielded a learning environment assessment scale with 28 items spanning seven domains [2]. The goal of this study was to determine the validity of the JHLES in Malaysia, and to ascertain its utility in detecting variation across medical school learning environments. We selected Malaysia because its rapid growth in medical schools (three accredited schools in 2000, and 21 accredited schools and up to 30 in operation today [3,4])—often in partnership with a foreign school—has uncertain implications for learning environment quality and makes the valid measurement of learning environments imperative. We studied pre-clerkship learning environments at three medical schools, each of which represents one of the three models of undergraduate medical education currently operating in Malaysia. Two schools were new international partnerships in Malaysia: Perdana University Graduate School of Medicine (PUGSOM), which is a graduate-entry program established with JHUSOM, and Perdana University-Royal College of Surgeons in Ireland School of Medicine (PURCSI), which is a school leaver program run by the Royal College of Surgeons in Ireland. The third medical school, Cyberjaya University College of Medical Sciences (CUCMS), is a private school-leaver program that was started in Malaysia without an international partner.

METHODS

Study design, subjects, and setting

This study involved the analysis of a cross-sectional survey given to all first- and second-year medical students at three medical schools at the end of the academic year. The characteristics of the three medical schools are summarized in Table 1. The students at CUCMS were surveyed at the end of the 2011-2012 academic year, and the students at PUGSOM and PURCSI were surveyed at the end of the 2012-2013 academic year. The surveys were administered electronically at PUGSOM and PURCSI, and given on paper, using optical mark reader sheets, at CUCMS. The responses were anonymous and the data were de-identified and analyzed by one author, who had no role in teaching or evaluating medical students at any of the schools.

Ethical approval

Institutional review board approval was obtained from Perdana University in Malaysia.

Survey composition

The student questionnaire was developed by members of our study team from institutions in the USA and Malaysia with extensive experience in medical education and educational research. All questionnaires included the JHLES [2], the Dundee Ready Education Environment Measure (DREEM) [5], a personal growth scale [6], several questions yielding an overall assessment of the learning environment, and questions about demographics.

The JHLES was developed over a series of iterative steps from 2010 to 2012 at JHUSOM [2]. Briefly, a 2010 survey of fourth-year JHUSOM students assessed events that impacted their perceptions of the learning environment [7]. Those results informed follow-up surveys of all JHUSOM students in 2011 and 2012; more than 30 items were used for exploratory factor analysis, which led to the final scale that cumulatively accounted for 57% of variance. The JHLES has 28 items, each with five-point response options, spanning six domains: community of peers (six items, 14% variance, α=0.91), faculty relationships (six items, 12% variance, α=0.80), academic climate (five items, 9% variance, α=0.86), meaningful engagement (four items, 9% variance, α=0.82), mentorship (two items, 5% variance, α=0.74), acceptance and safety (three items, 4% variance, α=0.58), and physical space (two items, 4% variance, α=0.66). Each item is scored from 1 to 5, meaning that total scores on the JHLES can range from 28 to 140. In February 2012, prior to survey administration in Malaysia, the JHLES was modified to remove language that was specific to JHUSOM in Baltimore, Maryland, and was piloted on paper with focus groups of preclinical students from PURCSI. To provide response processvalidity, the survey was discussed with these students to ensure that the meaning of items was clear and that their understanding of the learning environment was consistent with that theorized in previous research. This process led to only minor wording changes to items. For example, the JHLES item “The School of Medicine encourages ‘scholarship’ and innovation” was altered to read “The School of Medicine encourages ‘scholarly learning’ and innovation,” because students commonly thought of scholarship as tuition support.

The DREEM is a 50-item survey in which students respond with their level of agreement across a five-point scale. Items were grouped by its developers into five categories: perceptions of teachers, perceptions of teaching, academic self-perception, perceptions of atmosphere, and social self-perception. Each item is scored from 0 to 4, such that composite DREEM scores can range from 0 to 200.

The personal growth scale was originally a nine-item survey developed on internal medicine residents that asked about changes in attitudes and behaviors over time along a five-point Likert scale (1, much worse; 2, worse; 3, no change; 4, better; 5, much better). We altered the survey slightly to make it applicable to preclinical medical students and by framing questions about changes in their personal growth in relation to when respondents started medical school. This resulted in a 7-item scale. We scored items from 1 to 5, such that the total scores ranged from 7 to 35.

We also included three single-item global learning environment assessment variables that we have used in prior studies [2,9]. We asked students to rate their overall perception of the learning environment as (1) terrible, (2) poor, (3) fair, (4) good, or (5) exceptional, and their agreement with two statements: “The overall quality of the educational experience at the School of Medicine is excellent,” and “Based on my sense of the learning environment at the School of Medicine, I would recommend it to a close friend as a great school to attend.”

Statistical analysis

The basic descriptive statistics for the study population were tabulated with tests for significant differences applied as appropriate. The DREEM and JHLES total scores and domain scores were compared across schools using bivariate analysis with analysis of variation and the Kruskall-Wallis test, as appropriate. Multivariate analysis adjusted for students’ year in medical school, gender, and ethnicity. For institutional pairwise comparisons, we used t-tests with the Bonferroni correction. In order to establish relations to other variables’ validity for JHLES, we used Pearson correlation coefficients to determine associations among JHLES total scores, DREEM total scores, and personal growth scale total scores. Spearman correlation coefficients were calculated for associations between learning environment scale totals and global learning environment assessment variables. The internal structure of the JHLES was assessed by calculating values of Cronbach’s alpha to establish internal reliability and by calculating corrected item-total correlations. Stata ver. 13 (Stata Co., College Station, TX, USA) was used for all data analysis. Significance was set at P<0.05.

RESULTS

The overall response rate was 369/429 (86%). Complete surveys were collected from 51/55 students (93%) at PUGSOM, 97/134 students (72%) at PU-RCSI, and 221/240 students (92%) at CUCMS (Table 1). The average age was 25.1 years (standard deviation [SD]=1.6 years) at PUGSOM, 20.6 (SD=0.7) years at PURCSI, and 21.0 years (SD=0.9) at CUCMS, which were significantly different (P<0.001). Men numbered 16 (32%) at PUGSOM, 38 (40%) at PURCSI, and 82 (39%) at CUCMS; these differences were not statistically significant (P=0.72). Overall, 155 (42%) students were in year 1 and 213 (58%) were in year 2, which did not vary significantly across the three institutions (P=0.051). The ethnic make-up varied across the institutions (P=0.0001). The largest proportions of students were ethnically Chinese at PUGSOM (62%) and PURCSI (41%), and at CUCMS, most (74%) were Malay (Table 1).

Learning environment differences as reflected by the JHLES, DREEM, and global learning environment assessment variables

In comparisons across all three schools, no differences were detected in the total JHLES scores, but statistically significant differences were found between schools in five of seven (71%) JHLES domains, with four (57%) remaining significant after adjusting for medical school year and student gender and ethnicity (all P<0.05) (Table 2). PUGSOM had the highest ratings for ‘faculty relationships’ and ‘acceptance and safety,’ while CUCMS had the highest for ‘mentorship.’ The schools had nearly identical ratings for ‘academic climate’ (Table 2). No differences were seen in the total DREEM scores. Significant differences were seen in two of five (40%) DREEM categories, with one (20%) remaining significant after multivariate adjustment (Table 2). Each single-item global learning environment assessment variable failed to detect differences in the learning environments across the three schools (Table 3).

Relationship to other variables validity

The total JHLES score was highly correlated with the total DREEM score (r=0.78) across schools, with little difference at each school (PUGSOM, r=0.82; PURCSI, r=0.82; CUCMS, r=0.78).

The JHLES total was moderately to highly correlated with the three single-item global learning environment assessment variables and with the personal growth scale total (Table 4). The JHLES was more strongly correlated than DREEM to two of the three single-item global assessment variables and to the total score of the personal growth scale.

Evidence for internal structure validity

The JHLES showed high internal consistency for its total across all schools and at each school (Table 5). Each of its seven domains had values of Cronbach’s alpha within acceptable limits across schools, albeit with some site-to-site variability.

The corrected item-total correlations for the JHLES showed that all but two items had correlation coefficients above the acceptable level of 0.30 (“I am concerned that students are mistreated at the SOM” and “I feel concerned at times for my personal safety at the SOM”).

DISCUSSION

In this study of students’ perceptions of the preclinical learning environments at three Malaysian medical schools, we found that the JHLES was able to create unique learning environment domain profiles for each school. The JHLES also demonstrated sufficient internal reliability and correlation with external variables across the various student populations that were studied.

A systematic review of learning environment assessment tools showed that most did not have evidence for content, response process, internal structure, or a relationship to other variables’ validity [1]. The DREEM is by far the most published instrument, but only content and internal structure validity evidence have been demonstrated for it. The JHLES was recently developed according to guidelines for scale development, including exploratory factor analysis. Before this study, it had already been validated for face, content, response process, internal structure, and relation to other variables validity at JHUSOM [2]. In this study, we further substantiated the content, response process, internal structure, and relation to other variables validity by studying a new medical student population at three distinct medical schools in Malaysia.

The JHLES not only discriminated differences between schools but created unique learning environment profiles for each institution, while maintaining similar levels of evidence of validity at each site. This suggests that the scale may be suitable for a range of educational settings and particularly useful in assuring the quality of learning environments at new schools in Malaysia. Institutional profiles provide the opportunity for benchmarking, which is becoming a preferred approach for schools to identify areas of relative strength and weakness [10]. Moreover, such information may provide opportunities for institutions to collaborate and learn from one another. For example, PUGSOM may seek advice from CUCMS regarding their methods for fostering a ‘community of peers.’

Few studies of medical school learning environments have provided multi-institutional comparisons. We identified two studies in which DREEM was used for multi-institution comparisons. In 2008, differences were observed in total DREEM scores and scores in three of its five categories between two schools in Saudi Arabia; a school that used problem-based learning had higher ratings than one that did not [11]. In 2004, the DREEM found that a Scottish medical school had a higher total score and ratings in every category than three other medical schools (two in Saudi Arabia and one in Yemen) that had not been using modern teaching techniques [12]. In contrast to these studies, the DREEM found differences in only one category in this study of three recently opened Malaysian medical schools. All three schools in this study use modern curricula and interactive teaching methods. The DREEM contains several items that ask about interactive teaching methods, while many similar items were not retained during JHLES development because they did not vary among students at JHUSOM. This may suggest that as teaching methods have advanced, with changes in the formal and informal structures of institutions, the DREEM may no longer be as effective at detecting differences as it used to be.

As in our previous research, we found the JHLES and DREEM to be strongly correlated, suggesting that they are measuring a similar construct [9]. Also consistent with previous work, we found the JHLES to have superior concurrent criterion validity than the DREEM. The JHLES was able to detect differences in four of seven domains, after adjusting for covariates, whereas at 50 items, DREEM discriminated differences in one of its five categories. A recent confirmatory factor analysis performed on the DREEM at one school in Malaysia [13] suggested that 33 DREEM items do not meaningfully add to its assessment of the learning environment. Given the limitations of the DREEM, our findings suggest that future learning environment studies might prefer to use the JHLES, if further validation studies demonstrate its strengths.

We surveyed the very first two classes of students at PUGSOM and PURCSI, which are international medical education collaborations that follow different educational models. Compared to recent reports of DREEM ratings in Malaysia [14-17], which have ranged from 118 to 133, and a DREEM score of 131 reported at an international partnership in Saudi Arabia [11], students at PUGSOM and PURCSI thought favorably of their learning environments. Uniform ratings in the JHLES academic climate domain suggest that PUGSOM and PURCSI have been successful in exporting and contextualizing curricula to local settings. Low ratings on the JHLES mentorship domain at PUGSOM and PURCSI may very well identify a shared challenge that might be faced early on in the maturation of medical school operations. The additional finding that both schools had relatively low ratings of physical space lends face validity to our data, as both share an interim campus.

Several limitations of this work should be considered. First, only preclinical students were surveyed, and it is unknown whether JHLES would perform as well on international students during their clinical years. Second, we conceptualized personal growth as an outcome that may be realized while being immersed in a positive learning environment; however our cross-sectional study design does not permit causal inference, and it is unknown whether student learning environment perceptions lead to improved learning or clinical outcomes. Finally, while the JHLES performed well at three medical schools in Malaysia and at JHUSOM where it was developed, future studies will be required to determine validity evidence using other student populations.

In conclusion, JHLES appears to have been able to discriminate between discrete learning environment factors across the three educational settings in our study, creating unique domain profiles for each school. The interpretation of these profiles may allow medical schools to understand how they are currently supporting trainees and to identify areas needing attention.

Acknowledgments

SUPPLEMENTARY MATERIAL

References

• 1. Colbert-Getz JM, Kim S, Goode VH, Shochet R, Wright SM. Assessing medical students’ and residents’ perceptions of the learning environment: exploring validity evidence for the interpretation of scores from existing tools. Acad Med 2014;85:1687-1693. http://dx.doi.org/10.1097/ACM.0000000000000433 Article
• 2. Shochet RB, Colbert-Getz JM, Wright SM. The Johns Hopkins Learning Environment Scale: measuring medical student perceptions of the processes supporting professional growth. Acad Med 2015;90:810-818. http://dx.doi.org/10.1097/ACM.0000000000000706 ArticlePubMed
• 3. World Directory of Medical Schools [internet] Ferney-Voltaire: World Directory of Medical Schools; [Cited 2015 Feb 13]. Available from: http://www.wdoms.org/
• 4. Wikipedia. List of Malaysian medical schools [internet] San Francisco (CA): Wikipedia; [Cited 2015 Feb 13]. Available from: http://en.wikipedia.org/wiki/List_of_medical_schools_in_Malaysia
• 5. Roff S, Mcaleer S, Harden RM, Al-Qahtani M, Ahmed AU, Deza H, Groenen G, Primparyon P. Development and validation of the Dundee Ready Education Environment Measure (DREEM). Med Teach 1997;19:295-299. Article
• 6. Wright SM, Levine RB, Beasley B, Haidet P, Gress TW, Caccamese S, Brady D, Marwaha A, Kern DE. Personal growth and its correlates during residency training. Med Educ 2006;40:737-745. http://dx.doi.org/10.1111/j.1365-2929.2006.02499.x ArticlePubMed
• 7. Shochet RB, Colbert-Getz JM, Levine RB, Wright SM. Gauging events that influence students’ perceptions of the medical school learning environment: findings from one institution. Acad Med 2013;88:246-252. http://dx.doi.org/10.1097/ACM.0b013e31827bfa14 ArticlePubMed
• 8. Wiener CM, Thomas PA, Goodspeed E, Valle D, Nichols DG. “Genes to society”: the logic and process of the new curriculum for the Johns Hopkins University School of Medicine. Acad Med 2010;85:498-506. http://dx.doi.org/10.1097/ACM.0b013e3181ccbebf ArticlePubMedPMC
• 9. Tackett S, Shochet R, Colbert-Getz J, Rampal K, Shilkofski N, Wright S. Learning environment assessments of a single curriculum being taught at two medical Schools 10,000 miles apart. BMC Med Educ 2015;15:105. http://dx.doi.org/10.1186/s12909-015-0388-0 ArticlePubMed
• 10. Kirch DG, Prescott JE. From Rankings to Mission. Acad Med 2013;88:1064-1066. http://dx.doi.org/10.1097/ACM.0b013e318299efd9 Article
• 11. Zawawi AH, Elzubeir M. Using DREEM to compare graduating students’ perceptions of learning environments at medical schools adopting contrasting educational strategies. Med Teach 2012;34 Suppl 1:S25-31. http://dx.doi.org/10.3109/0142159X.2012.656747 ArticlePubMed
• 12. Al-Hazimi A, Zaini R, Al-Hyiani A, Hassan N, Gunaid A, Ponnamperuma G, Karunathilake I, Roff S, Mcaleer S, Davis MH. Educational environment in traditional and innovative medical schools: a study in four undergraduate medical schools. Educ Health (Abingdon) 2004;17:192-203.
• 13. Yusoff MSB. The Dundee ready educational environment measure: a confirmatory factor analysis in a sample of Malaysian medical students. Int J Hum and Soc Sci 2012;16:313-321. PubMedPMC
• 14. Al-Naggar RA, Abdulghani M, Osman MT, Al-Kubaisy W, Daher AM, Nor Aripin KN, Assabri A, Al-Hidabi DA, Ibrahim MI, Al-Rofaai A. The Malaysia DREEM: perceptions of medical students about the learning environment in a medical school in Malaysia. Adv Med Educ Pract 2014;5:177-184. http://dx.doi.org/10.2147/AMEP.S61805
• 15. Arzuman H, Yusoff MS, Chit SP. Big sib students’ perceptions of the educational environment at the School of Medical Sciences, Universiti Sains Malaysia, using Dundee Ready Educational Environment Measure (DREEM) Inventory. Malays J Med Sci 2010;17:40-47. PubMed
• 16. Lai N, Nalliah S, Jutti RC, Hla Y, Lim VK. The educational environment and self-perceived clinical competence of senior medical students in a Malaysian medical school. Educ Health (Abingdon) 2009;22:148.Article
• 17. Yusoff MSB, Jaa’far R, Arzuman H, Arifin WN, Mat Pa MN. Perceptions of medical students regarding educational climate at different phases of medical training in a Malaysian medical school. Educ Med J 2013;5:e30-41. http://dx.doi.org/10.5959/eimj.v5i3.146 ArticlePubMed

Figure & Data

References

Citations

Citations to this article as recorded by  

• A scoping review of the questionnaires used for the assessment of the perception of undergraduate students of the learning environment in healthcare professions education programs
  Banan Mukhalalati, Ola Yakti, Sara Elshami
  Advances in Health Sciences Education.2024; 29(4): 1501.     CrossRef
• Validation of the Polish version of the Johns Hopkins Learning Environment Scale–a confirmatory factor analysis
  Dorota Wójcik, Leszek Szalewski, Adam Bęben, Iwona Ordyniec-Kwaśnica, Robert B. Shochet
  Scientific Reports.2024;[Epub]     CrossRef
• Exploring medical students' experience of the learning environment: a mixed methods study in Saudi medical college
  Mohammed Almansour, Noura Abouammoh, Reem Bin Idris, Omar Abdullatif Alsuliman, Renad Abdulrahman Alhomaidi, Mohammed Hamad Alhumud, Hani A. Alghamdi
  BMC Medical Education.2024;[Epub]     CrossRef
• A multicenter cross-sectional study in China revealing the intrinsic relationship between medical students’ grade and their perceptions of the learning environment
  Runzhi Huang, Weijin Qian, Sujie Xie, Mei Cheng, Meiqiong Gong, Shuyuan Xian, Minghao Jin, Mengyi Zhang, Jieling Tang, Bingnan Lu, Yiting Yang, Zhenglin Liu, Mingyu Qu, Haonan Ma, Xinru Wu, Huabin Yin, Xiaonan Wang, Xin Liu, Yue Wang, Wenfang Chen, Min Li
  BMC Medical Education.2024;[Epub]     CrossRef
• Learning environment and its relationship with quality of life and burn-out among undergraduate medical students in Pakistan: a cross-sectional study
  Saadia Shahzad, Gohar Wajid
  BMJ Open.2024; 14(8): e080440.     CrossRef
• Validation of the Polish version of the DREEM questionnaire – a confirmatory factor analysis
  Dorota Wójcik, Leszek Szalewski, Adam Bęben, Iwona Ordyniec-Kwaśnica, Sue Roff
  BMC Medical Education.2023;[Epub]     CrossRef
• Association between patient care ownership and personal or environmental factors among medical trainees: a multicenter cross-sectional study
  Hirohisa Fujikawa, Daisuke Son, Takuya Aoki, Masato Eto
  BMC Medical Education.2022;[Epub]     CrossRef
• Measuring Students’ Perceptions of the Medical School Learning Environment: Translation, Transcultural Adaptation, and Validation of 2 Instruments to the Brazilian Portuguese Language
  Rodolfo F Damiano, Aline O Furtado, Betina N da Silva, Oscarina da S Ezequiel, Alessandra LG Lucchetti, Lisabeth F DiLalla, Sean Tackett, Robert B Shochet, Giancarlo Lucchetti
  Journal of Medical Education and Curricular Development.2020;[Epub]     CrossRef
• Developing an Introductory Radiology Clerkship at Perdana University Graduate School of Medicine in Kuala Lumpur, Malaysia
  Sarah Wallace Cater, Lakshmi Krishnan, Lars Grimm, Brian Garibaldi, Isabel Green
  Health Professions Education.2017; 3(2): 113.     CrossRef
• Trainers' perception of the learning environment and student competency: A qualitative investigation of midwifery and anesthesia training programs in Ethiopia
  Sharon Kibwana, Rachel Haws, Adrienne Kols, Firew Ayalew, Young-Mi Kim, Jos van Roosmalen, Jelle Stekelenburg
  Nurse Education Today.2017; 55: 5.     CrossRef