Skip to main content
Download PDF

Job preferences of master of public health students in China: a discrete choice experiment

BMC Medical Education volume 24, Article number: 24 (2024) Cite this article

Abstract

Background

The shortage of public health personnel and the uneven distribution between urban and rural areas are thorny issues in China. Master of public health (MPH) is an integral part of public health human resources in the future, and it is of far-reaching significance to discuss their work preferences. The present study wants to investigate the job preference of MPH, understand the relative importance of different job attributes, and then put forward targeted incentive measures.

Methods

Discrete choice experiment (DCE) was used to evaluate the job preference of MPHs in two medical colleges in Liaoning Province. Attributes include employment location, bianzhi, working environment, career development prospects, work value and monthly income. Thirty-six choice sets were developed using a fractional factorial design. Mixed logit models were used to analysis the DCE data.

Results

The final sample comprised 327 MPHs. All the attributes and levels included in the study are statistically significant. Monthly income is the most important factor for MPHs. For non-economic factors, they value career development prospects most, followed by the employment location. Respondents’ preferences are heterogeneous and influenced by individual characteristics. Subgroup analysis showed that respondents from different family backgrounds have different job preferences. Policy simulation suggested that respondents were most sensitive to a salary increase, and the combination of several non-economic factors can also achieve the same effect.

Conclusions

Economic factors and non-economic factors significantly affect the job preference of MPHs. To alleviate the shortage and uneven distribution of public health personnel, more effective policy intervention should comprehensively consider the incentive measures of the work itself and pay attention to the individual characteristics and family backgrounds of the target object.

Peer Review reports

Background

Since the beginning of this century, the challenges in the public health field have become more severe, and the forms have significantly changed [1]. The outbreak of novel coronavirus at the end of 2019 has spread in more than 200 countries worldwide, causing severe damage to human health, and even causing the whole society to shut down temporarily. Facing public health emergencies, the public health workforce with rich theoretical knowledge and practical abilities plays a crucial role in the health system [2]. To better deal with complex and changeable public health events, it is a crucial step to improve public health education [3]. According to the national health service system planning outline (2015–2020), by 2020, China would have 0.83 public health personnel per 1,000 permanent residents. However, in 2020, this figure was only 0.66. The shortage of public health personnel is indisputable in China [4].

The shortage of health workforce harms the construction of the medical and health system [5]. To promote the development of public health, the Ministry of Education of China launched the full-time Master of Public Health (MPH) postgraduate program in 2009 [6]. According to the China Health Statistics Yearbook in 2018, in terms of educational structure, 54% of the staff in the centres for disease control and prevention at all levels in China have only a junior college degree, 37% had a bachelor’s degree and only 7% had a master’s degree. However, among the public health students worldwide, 37% were undergraduates, 49% were masters, and 14% were doctoral students [7]. Regarding the proportion of highly educated public health personnel, China has not reached the average level in the world. Since the COVID-19 outbreak, China has significantly increased enrollment in MPH.

The uneven distribution of health workers in urban and rural areas is also a concern directly related to people’s access to basic health services [8]. In many countries, there is an uneven distribution of health workers in urban and rural areas or remote areas, and coordinating the balanced health workforce distribution has become a key policy issue [9, 10]. In China, for example, the number of health workers per unit population in urban areas is much higher than in rural areas, and the gap is still widening [11]. The Dakar region, which accounts for only 23% of Senegal’s total population, has more than 60% of the country’s medical staff [12]. About half of the world’s population lives in rural and remote areas, but only a quarter of doctors and less than a third of nurses serve this half [13]. In addition, the WHO report predicts that, over the next few decades, about 40% health professionals worldwide will leave their positions due to little incentives and low wages [14]. Therefore, it is increasingly urgent to formulate reasonable policies and measures to improve the attractiveness of job prospects in rural areas and promote the balanced allocation of urban and rural health human resources.

The public health system plays a vital role in developing China’s health sector and protecting people’s health [15]. Public health is neither solely a hard nor social science and graduates majoring in public health have relatively flexible employment [16]. MPHs have the ability to adjust the focus of work in time according to different environments and can effectively cooperate with other members of the team to realize the complementarity of different majors and improve work efficiency [7]. The job preference of students majoring in health will affect the geographical distribution of health workers in the future [17]. To better allocate public health personnel, it would be a wise choice to investigate the job preference of MPH.

Discrete choice experiment (DCE) can quantitatively measure individual choice preferences [18] by creating simulated choice scenarios [19]. The DCE provides a weighed relevance to the attributes to distinguish highly valued ones [17]. Therefore, quantitative information on the relative strength of selected attributes can be determined, and trade-offs between these attributes and the probability of individuals taking up these jobs can be ascertained [20, 21]. The strength of DCE lies in its ability to assess various scenarios that may not be observed in real-world situations, which makes DCE more widely used in the field of health. For example, scholars in China have studied the employment preferences of pharmaceutical students [22]. In addition, the research on the work preference of community health workers [23] and medical staff [24] has already existed. However, up to now, there is no research on MPHs in China or the world.

In this study, MPHs from two medical universities in Northeast China were taken as the research subjects. The attributes and relative importance of their work preferences were analysed through DCEs. We believe that our findings can provide a basis for public health policymakers to address the unbalanced distribution of human resources in public health and the increasingly severe brain drain phenomenon and improve the construction of China’s health policy systems and mechanisms.

Methods

Study context

This study was conducted in Liaoning, a province located in northeast China. Regarding the absolute number of health workers, Liaoning has the most significant number of health workers among the three northern provinces (Liaoning, Jilin, and Heilongjiang) in China. China Health Statistics Yearbook (2020) shows that the number of public health physicians per 10,000 population in Liaoning Province is 0.82, slightly lower than the national average of 0.83.

Considering the number and representativeness of the respondents, we chose the two largest local medical colleges—China Medical University and Dalian Medical University. MPHs, who have yet to graduate, are our target research subjects. Convenience sampling is adopted in this study, and an anonymous web-based survey was conducted through the Wenjuanxing platform (one of the most commonly used online questionnaire survey platforms in China) from 8 December 2021 to 2 January 2022.

Previous studies have shown that a small sample size is a significant advantage of DCE. In most cases, a sample size greater than 100 can provide sufficient support for the subsequent analysis of preference data [25]. According to the rule of thumb put forward by Johnson and Orme [26, 27], the sample size of DCE depends on three factors: the maximum number of levels in any attribute (c); the number of DCE questions in each questionnaire (t); the number of options contained in each DCE question (a). The calculation formula is as follows:

$$ N>\frac{500\times c}{t\times a}$$

According to the rule of thumb, the minimum number of respondents required for this study was 63.

Study design

DCE can design a hypothetical working situation with two options for respondents to choose and determine the relative importance of attributes according to respondents’ choices [28, 29]. The design and analysis of the experiment are in accordance with the user guide of DCE published by the WHO [20] and the instruction manual written by Lancsar [30].

Identification of attributes and levels

DCE is the most powerful tool to measure individual choice behavior [31]. In a DCE, there are several choice sets, and each choice set is described by specific attributes and levels [32]. The development of attributes and levels is the first and most critical step in DCE experimental design [33]. Through literature review and consideration of the characteristics of public health specialty, we preliminarily determined eight attributes [22, 34,35,36,37,38], including employment location, bianzhi, working pressure, working environment, career development prospect, monthly income, training opportunity, and professional compliance. Then, we consulted experts in related majors, and they suggested deleting the attribute of training opportunities because the career development prospect includes training opportunities to some extent. Then, two focus groups were conducted among MPH from China Medical University. They needed to discuss the attributes they would pay attention to when choosing a job. In addition, they also needed to discuss the seven attributes and levels that had been initially determined until they finally reached a consensus. According to their feedback, the attributes of working pressure and professional compliance were deleted, and the attribute of work value was added. Prior to the data collection phase, this study was pre-tested in the MPH of China Medical University, and some minor modifications were made. Finally, six attributes were included in the study, as shown in Table 1. A monetary attribute (monthly income) is always included in DCE to estimate the respondents’ willingness to pay for various attributes [29].

Table 1 Attributes and levels used in discrete choice experiment
Full size table

Experimental design

It is helpful to realize unbiased statistical response efficiency by following the standard approaches of DCE design [39]. This study contains six attributes, two of which have a level of 2, and the other four have a level of 3. A full factorial design will produce 324 (22 × 34) potential combinations and 52326 ((324 × 323)÷2) potential choice tasks. The number of potential scenarios presented to the respondents was reduced by a fractional factorial experiment design, and the ‘dcreate’ module of STATA 15.0 was used to optimize the D-efficiency, minimize the overlap among attributes levels, and maximize level balance and orthogonality. The design finally contains 36 choice sets. Studies have shown that the average number of choice sets implemented by DCE for health workers in the past was controlled at about 12 [40]. For reducing the burden of the respondents, the 36 sets were further divided into three blocks, each containing 12 choice sets. An opt-out was included in the second-stage question after each DCE task to allow for unconditional choices [41]. We used an unlabeled DCE for improving the reliability of job preference estimation [42]. Table 2 presents an example of a complete DCE choice set. Respondents need to choose the one they prefer from two hypothetical jobs, and then make a choice whether to choose the job in real life [41]. All three blocks have a duplicate selection set to check the internal consistency, and all respondents were randomized to receive one of the 3 blocks.

Table 2 An example of a choice pair
Full size table

Data collection

In order to ensure the quality of the research, before the formal investigation, we conducted a small-scale preliminary investigation (n = 58) on MPH of China Medical University. In the pre-survey, we tested the questionnaire’s comprehensibility, acceptability and validity. Then, we adjusted the language of the questionnaire and the layout of the questions according to the pre-survey results.

The formal questionnaire consists of three parts: preface (purpose of investigation), demographic information survey and job preference survey. Before the investigation, participants were well-informed about the purpose and protocol of the study, and informed consent was obtained from them. This study was approved by the Ethics Committee of China Medical University.

Statistical analysis

We use Stata (version 15.0) for data analysis. χ2 test was adopted to compare category variables in descriptive statistical analysis. DCE data were analysed through mixed logit model, as it has smaller value of the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) than conditional logit model. The smaller the value of the AIC or BIC, the more accurate and reasonable the corresponding regression model in the context of research [43, 44]. All attributes were specified as having a random component. All characteristics were measured as a categorical variable with the first level taken as the reference level, except that the monthly income characteristic was analyzed as a continuous variable. Furthermore, DCEs are based on the random utility theory (RUT) [45]. It assumes that respondent n will choose alternative j in choice scenario c if that alternative provides the most satisfaction out of all other alternatives [46]. It can be specified as:

Unjc = (β0 + η0n) ASCnjc.

+ (β1 + η1n) Employment location City.

+ (β2 + η2n) BianzhiYes.

+ (β3 + η3n) Working environment Ordinary.

+ (β4 + η4n) Working environment Relatively good.

+ (β5 + η5n) Career development prospects Ordinary.

+ (β6 + η6n) Career development prospects Relatively good.

+ (β7 + η7n) Work value Ordinary.

+ (β8 + η8n) Work value High value.

+ (β9 + η9n) Monthly income + εnjc.

Where Unjc is the utility of individual n from choosing alternative j in choice scenario c; β is a vector of coefficients reflecting the desirability of the attributes; η reflects the degree of heterogeneity among respondents; ASCnjc is the alternative-specific constant; and εnjc signifies the unobservable random component. An advantage of the mixed logit model relaxes the irrelevant alternatives (IIA) assumption by allowing coefficients to vary between individuals, thus it can consider the potential heterogeneity of respondents’ preferences [47]. Furthermore, in this model the desirability of attributes constitutes a vector of average preferences of the population for each attribute (β) and the individual’s specific preference components (η) [48]. Both the mean and standard deviation (SD) of β will be estimated, from which preference heterogeneity can be assessed [49].

When there are economic attributes (such as monthly income) in each set research attribute, by calculating the ratio of regression coefficients of economic attributes and non-economic attributes we can get the evaluation of the monetary value of each non-economic attribute (i.e. willingness to pay) of the respondents. The willingness to pay (WTP) can measure the income that the respondents are willing to give up in order to get a certain attribute or the compensation income expected by accepting a certain attribute [50]. A simulation study was conducted to predict that the uptake rates of MPH for rural versus city jobs change as the levels of job attributes are changed.

Result

A total of 401 MPHs filled out the questionnaires through the wenjuanxing platform. Thirty-four (8.5%) respondents were rejected because of their short answer time (an answer time of less than 50% of the median answer time will be defined as answering too fast) [51]. There are also 40 (10.0%) respondents who were excluded because they failed the internal consistency test. In the end, the total number of valid questionnaires in this study was 327.

Characteristics of the respondents

Table 3 presents the basic information of the respondents included in the analysis. The average age of the MPH (n = 327) was 24.2 ± 1.8 years. Female (77.68%) accounts for the overwhelming majority of participants, which was consistent with a study on the employment intention of Chinese students majoring in public health [52]. There are 191 (58.41%) people from non-only-child families, which is slightly higher than those from only-child families. Most were grade 1 master respondents (48.62%) and from rural areas (43.73%). About 78.59% of the students plan to work in public health and preventive medicine after graduation, and about half are neutral about their future employment forms. See Table 3 for more details.

Table 3 Characteristics of the respondents (n = 327)
Full size table

Mixed logit estimates for MPH’s job preference

The results of the regression analysis of the mixed logit model are shown in Table 4. Among the six attributes included in the study, at least one level of each attribute is statistically significant, which means that all these attributes are crucial to the job preference of MPH. The sign of each regression coefficient implies that the respondents have obtained a higher utility level from a higher attribute level and made a rational choice [53]. In choosing a job, MPH considers not only economic factors but also non-economic factors. They pay special attention to their career development prospects (β = 1.046, p < 0.001) and strongly prefer jobs with relatively good career development prospects rather than relatively poor ones. The employment location (β = 1.019, p < 0.001) is also their key attribute, and they tend to seek a job in the city rather than in the country. The alternative specific constant (ASC) for the opt-out is significantly negative, which indicates that the respondents are more likely to express their preference for choosing a job, regardless of the attributes and level presented [54], that is, they show a positive attitude towards choosing a job. For monthly income, employment location, bianzhi, working environment and relatively good career development prospects, the SDs of those random coefficients were statistically significant, which indicated preference heterogeneity was present for those attribute levels. The relative importance scores of each attribute are calculated and presented in Fig. 1. As we can see from the Figure, the monthly income and career development prospects dominate the choice of a job for MPH. Work value is regarded as the least important attribute, with a score of 0.074.

Table 4 Mixed logit estimates for MPH’s job preference
Full size table
Fig. 1
figure 1

The relative importance scores of each attribute. Higher scores indicate that the attribute is more important to the respondents

\(({\rm{the}}\,{\rm{relative}}\,{\rm{importance}}\,{\rm{scores}}\,{\rm{of}}\,{\rm{attribute}}\,i = \frac{{{\rm{Maximum}}\,{\rm{utility}}\,{\rm{of}}\,{\rm{attribute}}\,i}}{{{\rm{Total}}\,{\rm{utility}}}})\)

Full size image

Willingness to pay for job attributes

The results of WTP estimates are documented in Table 5. The most crucial job attributes, ranked in terms of willingness to sacrifice monthly income, are career development prospects, employment location and bianzhi. They were willing to pay RMB 1246.3 (US$ 184.6) monthly income for a job with relatively good career development prospects than a job with relatively poor career development prospects. In order to leave the country to work in the city, they are willing to give up RMB 1214.5 (US$ 179.9) per month.

Table 5 Estimated willingness to pay (WTP) for job attributes among MPH
Full size table

Subgroup analysis

In view of our analysis of the existing literature and the results of focus group interviews, this study carried out a subgroup analysis on the respondents based on whether they were only child, where their families belonged, and the total annual income of their families. We found that there were differences between the job preference and the willingness to pay of related attributes of respondents with different individual characteristics. Respondents from only child families pay the most attention to the employment location when choosing jobs. It is worth noting that respondents whose families belong to cities and counties and whose total annual income is more than RMB 100,000 (US$ 14821.8) all show the same preference. In addition, the most crucial job attribute for respondents whose total annual household income is less than RMB 100,000 is bianzhi. They are willing to pay RMB 1169.3 (US$ 173.3) for a job with bianzhi. More details are provided in Additional file 1: Supplementary Tables 1–3. Considering the difficulty and efficiency of model fitting, basic demographic variables were not included in our subgroup analysis. This will be further discussed in the follow-up study.

Changes in employment rate under different policy interventions

Figure 2 shows the changes in the probability of taking a job in country. It can be seen from the figure that under the benchmark condition (Monthly income: RMB 3500, Bianzhi: No, Working environment: Relatively poor, Career development prospects: Relatively poor, Work value: Low value), the probability of MPHs choosing to work in counties and towns is only 10.1%. Holding all else the same, if monthly income increased from RMB 3500 to RMB 5000, the probability of choosing a country job would increase to 55.7%. When the three non-economic factors, such as employment location, bianzhi and career development prospects, change at the same time (9: 2 + 3 + 4), the probability of MPH choosing to work in counties increases to 90.5%, slightly higher than the probability of working in cities under the benchmark conditions.

Fig. 2
figure 2

The changes in the probability of taking a job in country. With the improvement of working attributes, the probability of MPH willing to work in remote areas changes. Each serial number represents the corresponding attribute, and the combination of serial numbers represents the combination of different attributes (Baseline: Monthly income_RMB 3500, Bianzhi_No, Working environment_Relatively poor, Career development prospects_Relatively poor, Work value_Low value)

Full size image

Discussion

The present study was the first DCE study conducted among MPHs to investigate their preferences in job attributes. DCE is helpful as it can check the respondents’ declarative preference and willingness to choose rather than their explicit preference (actual choice) [55]. In this study, we finally determined six attributes closely related to MPH’s job preference. Mixed Logit analysis showed that all the attributes and levels were statistically significant, which is consistent with other studies [18, 24]. Monetary attribute has the most significant influence, consistent with the results of research on job preference conducted in many other places [23, 37, 47]. The employment location is one of the most crucial nonmonetary attributes of MPH. There is a big gap between rural and urban areas in terms of infrastructure and educational opportunities for children, which may be one of the reasons for the shortage of rural health human resources [56]. Similar results have been reported in other human resource DCE studies [22]. Lehmann et al. have also pointed out that students’ home location will have an impact on their future career choice [56]. Therefore, when formulating relevant policies to attract MPH to work in remote areas, the family background of the target object should be considered.

Bianzhi is a strong motivator of MPH. Researches on job preference in China generally includes this attribute and usually has a relatively significant influence [57, 58]. Chinese jokingly call bianzhi “iron rice bowl”, which means that the work with bianzhi has strong stability and will likely accrue better job security [59]. Chen surveyed employment intention in medical higher vocational colleges [60]. The study showed that students in areas with relatively backward economic and cultural development levels have higher requirements for job stability, and our research found a similar result. The most crucial job attribute for students from relatively poorer families is the bianzhi, while for students from relatively wealthier families is the employment location. For policymakers, in order to attract MPH to devote themselves to the development of rural health undertakings, it may be more effective to provide preparation for students from relatively low-income families.

As expected, the essential attribute for all the respondents is the economic factor. As we can see from the simulated working situation, by raising the monthly income level by one level (from RMB 3500 to RMB 5000), the respondents’ willingness to work in counties will increase by 45%. At present, China’s health service personnel are not satisfied with the salary, and improving income and welfare will have positive significance for retaining them [61, 62]. The monthly income level of workers is significantly related to the financial strength of the employment location [63]. Considering that the economic level and development level of counties are generally lower than those of cities, it is a more practical method to combine economic incentives with non-economic incentives.

Work value is the least valued attribute of MPH. According to Maslow’s Hierarchy of Needs [64], the desire of the previous level is satisfied, there is a chance to satisfy the desire of the next level. The pursuit of work value is a manifestation of self-realization, so it can be considered that its priority will be lower than other attributes. McAuliffe et al. pointed out in their research on the work preference of health workers providing obstetric care that once work reward can meet their basic needs, and other job attributes will become more important than salary [65]. Therefore, it can be inferred that the respondents have no confidence in the welfare of their future jobs. When making policies to attract MPH to relatively remote and backward areas, it is necessary to meet their basic living needs.

When formulating targeted recruitment policies, it is necessary to understand the individual characteristics of the respondents fully. Research has shown that gender has a significant impact on job selection [22, 66]. According to our research, respondents from only-child families prefer to work in cities and relatively good working environments. Two explanations are available. Firstly, the one-child families are more concentrated in urban areas [67]. Because of the influence of students’ geographical origin on their career choice [56], students from urban areas are more inclined to work in cities [34]. Secondly, the economic conditions of the only-child families are relatively better [68], and their children will be more inclined to choose places with better supporting facilities and more convenient life. Thus, more attention can be paid to the applicant’s family situation when recruiting MPH in underdeveloped counties and towns.

The research has the following three limitations. Firstly, the coverage of the sample is not wide enough and the representativeness is slightly insufficient. Secondly, this study focused on job seekers. For formulating more accurate and effective policies and measures, preference of the employer should also be considered. Thirdly, we did not explore differences in the results between the two institutions due to the limited sample size. However, the learning environment of the respondents often has an important impact on their career planning, and the differences should be explored in future studies.

Conclusions

The present study found that monthly income and employment location are two crucial job attributes in MPH’s career selection process. MPH with different individual characteristics has different preferences for job attributes. Combining economic and non-economic factors could be a more effective and feasible measure. Our results suggest a variety of possibilities to improve MPHs’ deployment in rural settings. The findings of this study will help policymakers to design a more effective recruitment plan for MPH in China, alleviate the shortage and uneven distribution of public health personnel, and thus construct a more reasonable public health system.

Data Availability

The datasets used or analysed during the current study are available from the corresponding author on reasonable request.

Abbreviations

AIC:

Akaike information criterion

BIC:

Bayesian information criterion

DCE:

Discrete choice experiment

MIXL:

Mixed logit model

MPH:

Master of Public Health

RUT:

Random utility theory

References

  1. Begg MD, Galea S, Bayer R, Walker JR, Fried LP. MPH education for the 21st century: design of Columbia University’s new public health curriculum. Am J Public Health. 2014;104(1):30–6.

    Article  Google Scholar 

  2. Cao Y, Shan J, Gong Z, Kuang J, Gao Y. Status and challenges of public health emergency management in China related to COVID-19. Front Public Health. 2020;8:250.

    Article  Google Scholar 

  3. Fried LP, Begg MD, Bayer R, Galea S. MPH education for the 21st century: motivation, rationale, and key principles for the new Columbia public health curriculum. Am J Public Health. 2014;104(1):23–30.

    Article  Google Scholar 

  4. Li C, Sun M, Wang Y, Luo L, Yu M, Zhang Y, et al. The centers for disease control and prevention system in China: trends from 2002–2012. Am J Public Health. 2016;106(12):2093–102.

    Article  Google Scholar 

  5. Guilbert JJ. The World Health Report 2006: working together for health. Educ Health (Abingdon). 2006;19(3):385–7.

    Article  Google Scholar 

  6. Ren T, Wang XW, Song J, Qin XY, Zhu YP, Wu YQ, et al. Investigation on the current status of the cultivation of the master of public health (MPH) in colleges in China. Zhonghua Yu Fang Yi Xue Za Zhi. 2019;53(4):419–20. (in Chinese).

    Google Scholar 

  7. Averill MM, Dillon-Sumner L, Stergachis A, Sconyers J, Summerside N, Brazg T, et al. Integrating public health students into interprofessional education. J Interprof Care. 2020;34(3):427–30.

    Article  Google Scholar 

  8. Rockers PC, Jaskiewicz W, Wurts L, Kruk ME, Mgomella GS, Ntalazi F, et al. Preferences for working in rural clinics among trainee health professionals in Uganda: a discrete choice experiment. BMC Health Serv Res. 2012;12:212.

    Article  Google Scholar 

  9. Buchan J, Couper ID, Tangcharoensathien V, Thepannya K, Jaskiewicz W, Perfilieva G, et al. Early implementation of WHO recommendations for the retention of health workers in remote and rural areas. Bull World Health Organ. 2013;91(11):834–40.

    Article  Google Scholar 

  10. Anand S, Fan VY, Zhang J, Zhang L, Ke Y, Dong Z, et al. China’s human resources for health: quantity, quality, and distribution. Lancet. 2008;372(9651):1774–81.

    Article  Google Scholar 

  11. Liu T, Li S, Yang R, Liu S, Chen G. Job preferences of undergraduate nursing students in eastern China: a discrete choice experiment. Hum Resour Health. 2019;17(1):1.

    Article  Google Scholar 

  12. Zurn P, Codjia L, Sall FL, Braichet JM. How to recruit and retain health workers in underserved areas: the Senegalese experience. Bull World Health Organ. 2010;88(5):386–9.

    Article  Google Scholar 

  13. Increasing access to health workers in remote and rural areas through improved retention: global policy recommendations. Geneva: World Health Organization; 2010.

    Google Scholar 

  14. Abate HK, Mekonnen CK. Job satisfaction and associated factors among health care professionals working in public health facilities in Ethiopia: a systematic review. J Multidiscip Healthc. 2021;14:821–30.

    Article  Google Scholar 

  15. Wang L, Wang Z, Ma Q, Fang G, Yang J. The development and reform of public health in China from 1949 to 2019. Global Health. 2019;15(1):45.

    Article  Google Scholar 

  16. Beck AJ, Leider JP, Krasna H, Resnick BA. Monetary and nonmonetary costs and benefits of a public health master’s degree in the 21st Century. Am J Public Health. 2020;110(7):978–85.

    Article  Google Scholar 

  17. Bao M, Huang C. Job preferences of medical and nursing students seeking employment in rural China: a discrete choice experiment. BMC Med Educ. 2021;21(1):146.

    Article  Google Scholar 

  18. Valet P, Sauer C, Tolsma J. Preferences for work arrangements: a discrete choice experiment. PLoS ONE. 2021;16(7):e0254483.

    Article  Google Scholar 

  19. Mangham LJ, Hanson K, McPake B. How to do (or not to do)… designing a discrete choice experiment for application in a low-income country. Health Policy Plan. 2009;24(2):151–8.

    Article  Google Scholar 

  20. Ryan M, Kolstad JR, Rockers PC, Dolea C. How to conduct a discrete choice experiment for health workforce recruitment and retention in remote and rural areas: a user guide with case studies (English). Washington, D.C.: World Bank Group.

  21. Miranda JJ, Diez-Canseco F, Lema C, Lescano AG, Lagarde M, Blaauw D, et al. Stated preferences of doctors for choosing a job in rural areas of Peru: a discrete choice experiment. PLoS ONE. 2012;7(12):e50567.

    Article  Google Scholar 

  22. Liu P, Liu S, Gong T, Li Q, Chen G, Li S. Job preferences of undergraduate pharmacy students in China: a discrete choice experiment. Hum Resour Health. 2021;19(1):79.

    Article  Google Scholar 

  23. Agarwal S, Abuya T, Kintu R, Mwanga D, Obadha M, Pandya S, et al. Understanding community health worker incentive preferences in Uganda using a discrete choice experiment. J Glob Health. 2021;11:07005.

    Article  Google Scholar 

  24. Angell B, Khan M, Islam R, Mandeville K, Naher N, Hutchinson E, et al. Incentivising doctor attendance in rural Bangladesh: a latent class analysis of a discrete choice experiment. BMJ Glob Health. 2021;6(7):e006001.

    Article  Google Scholar 

  25. de Bekker-Grob EW, Donkers B, Jonker MF, Stolk EA. Sample size requirements for discrete-choice experiments in healthcare: a practical guide. Patient. 2015;8(5):373–84.

    Article  Google Scholar 

  26. Orme B. Sawtooth software sample size issues for conjoint analysis studies. Research Paper Series 98382. 1998.

  27. Johnson R, Orme B. Getting the Most from CBC. Sawtooth Software research Paper Series 98382. 2003.

  28. Hasman A. Eliciting reasons: empirical methods in priority setting. Health Care Anal. 2003;11(1):41–58.

    Article  Google Scholar 

  29. Song K, Scott A, Sivey P, Meng Q. Improving Chinese primary care providers’ recruitment and retention: a discrete choice experiment. Health Policy Plan. 2015;30(1):68–77.

    Article  Google Scholar 

  30. Lancsar E, Louviere J. Conducting discrete choice experiments to inform healthcare decision making: a user’s guide. PharmacoEconomics. 2008;26(8):661–77.

    Article  Google Scholar 

  31. Louviere JJ, Hensher DA, Swait JD. Stated choice methods: analysis and application. Cambridge: Cambridge University; 2000.

    Book  Google Scholar 

  32. Holte JH, Kjaer T, Abelsen B, Olsen JA. The impact of pecuniary and non-pecuniary incentives for attracting young doctors to rural general practice. Soc Sci Med. 2015;128:1–9.

    Article  Google Scholar 

  33. Coast J, Horrocks S. Developing attributes and levels for discrete choice experiments using qualitative methods. J Health Serv Res Policy. 2007;12(1):25–30.

    Article  Google Scholar 

  34. Liu S, Li S, Yang R, Liu T, Chen G. Job preferences for medical students in China: a discrete choice experiment. Med (Baltim). 2018;97(38):e12358.

    Article  Google Scholar 

  35. Liu S, Li S, Li Y, Wang H, Zhao J, Chen G. Job preferences for healthcare administration students in China: a discrete choice experiment. PLoS ONE. 2019;14(1):e0211345.

    Article  Google Scholar 

  36. Wu P, Li Z, Guo W, Wang L, Chang X, Zhang Y, et al. Optimizing the practice environment for medical staff in the post-pandemic era: a discrete choice experiment. Front Public Health. 2022;10:911868.

    Article  Google Scholar 

  37. Efendi F, Chen CM, Nursalam N, Andriyani NW, Kurniati A, Nancarrow SA. How to attract health students to remote areas in Indonesia: a discrete choice experiment. Int J Health Plann Manage. 2016;31(4):430–45.

    Article  Google Scholar 

  38. Meng D, Xu G, He L, Zhang M, Padula WV, Davidson PM. Nursing students’ perceived value of the work environment: a discrete choice experiment. Geriatr Nurs. 2021;42(1):94–8.

    Article  Google Scholar 

  39. Reed Johnson F, Lancsar E, Marshall D, Kilambi V, Mühlbacher A, Regier DA, et al. Constructing experimental designs for discrete-choice experiments: report of the ISPOR conjoint analysis experimental design good research practices task force. Value Health. 2013;16(1):3–13.

    Article  Google Scholar 

  40. Mandeville KL, Lagarde M, Hanson K. The use of discrete choice experiments to inform health workforce policy: a systematic review. BMC Health Serv Res. 2014;14:367.

    Article  Google Scholar 

  41. Cheng J, Pullenayegum E, Marshall DA, Marshall JK, Thabane L. An empirical comparison of methods for analyzing correlated data from a discrete choice survey to elicit patient preference for colorectal cancer screening. BMC Med Res Methodol. 2012;12:15.

    Article  Google Scholar 

  42. de Bekker-Grob EW, Hol L, Donkers B, van Dam L, Habbema JD, van Leerdam ME, et al. Labeled versus unlabeled discrete choice experiments in health economics: an application to colorectal cancer screening. Value Health. 2010;13(2):315–23.

    Article  Google Scholar 

  43. Train K. Discrete choice methods with simulation. 2nd ed. Cambridge: Cambridge University; 2009.

    Google Scholar 

  44. Hole AR. Fitting mixed logit models by using maximum simulated likelihood. Stata J. 2007;7(3):388–401.

    Article  Google Scholar 

  45. Meijer E, Rouwendal J. Measuring welfare effects in models with random coefficients. J Appl Econ. 2006;21(2):227–44.

    Article  Google Scholar 

  46. McFadden D. Conditional logit analysis of qualitative choice behavior. In: Zarembka P, editor. Economic theory and mathematical economics. New York: Academic; 1974. pp. 105–42.

    Google Scholar 

  47. Dahlberg M, Eklöf M. Relaxing the IIA assumption in locational choice models: a comparison between conditional logit, mixed logit, and multinomial probit models. Working Paper. 2003; No. 2003:9.

  48. Hensher DA. Accounting for scale heterogeneity within and between pooled data sources. Transp Res Pt A-Policy Pract. 2012;46(3):480–6.

    Article  Google Scholar 

  49. Yin Y, Peng Q, Ma L, Dong Y, Sun Y, Xu S et al. QALY-type preference and willingness-to-pay among end-of-life patients with cancer treatments: a pilot study using discrete choice experiment. Qual Life Res. 2023.

  50. Park BH, Ko Y. Hospital preferences of nursing students in Korea: a discrete choice experiment approach. Hum Resour Health. 2016;14(1):58.

    Article  Google Scholar 

  51. Greszki R, Meyer M, Schoen H. Exploring the effects of removing too fast responses and respondents from web surveys. Public Opin Quart. 2015;79(2):471–503.

    Article  Google Scholar 

  52. Liang BW, Zhan YW, Feng AP, Gao YX, Li PY, Shu YL, et al. A survey on the willingness of college students majoring in public health to work in their majors. Zhonghua Liu Xing Bing Xue Za Zhi. 2021;42(3):562–8. (in Chinese).

    Google Scholar 

  53. Mangham LJ, Hanson K. Employment preferences of public sector nurses in Malawi: results from a discrete choice experiment. Trop Med Int Health. 2008;13(12):1433–41.

    Article  Google Scholar 

  54. Marshall HS, Chen G, Clarke M, Ratcliffe J. Adolescent, parent and societal preferences and willingness to pay for meningococcal B vaccine: a Discrete Choice Experiment. Vaccine. 2016;34(5):671–7.

    Article  Google Scholar 

  55. Lagarde M, Blaauw D. A review of the application and contribution of discrete choice experiments to inform human resources policy interventions. Hum Resour Health. 2009;7:62.

    Article  Google Scholar 

  56. Lehmann U, Dieleman M, Martineau T. Staffing remote rural areas in middle- and low-income countries: a literature review of attraction and retention. BMC Health Serv Res. 2008;8:19.

    Article  Google Scholar 

  57. Yan D, Zhang H, Chang J, Zhao Y, Liu X. Job preferences of doctors at township health center: a discrete choice experiment in three provinces of China. Chin J Health Policy. 2014;7(4):44–8. (in Chinese).

    Google Scholar 

  58. Zhang H, Pan W, Yan D, Hu D, Liu X. Job preferences of nurses at township health centers: based on a discrete choice experiment. Chin Health Resour. 2015;18(5):338–41. (in Chinese).

    Google Scholar 

  59. Zhang Y, Steege LM, Pavek KU, Brown RL, Zhang Y. Identifying patterns of occupational stress trajectories among newly graduated nurses: a longitudinal study. Int J Nurs Stud. 2019;99:103332.

    Article  Google Scholar 

  60. Chen YM. Investigation and analysis on employment intention of higher vocational college students in medical colleges. Health Voc Educ. 2017;35(6):124–6. (in Chinese).

    Google Scholar 

  61. Shi L, Song K, Rane S, Sun X, Li H, Meng Q. Factors associated with job satisfaction by Chinese primary care providers. Prim Health Care Res Dev. 2014;15(1):46–57.

    Article  Google Scholar 

  62. Wen T, Zhang Y, Wang X, Tang G. Factors influencing turnover intention among primary care doctors: a cross-sectional study in Chongqing, China. Hum Resour Health. 2018;16(1):10.

    Article  Google Scholar 

  63. Liu Y, Long C. Urban and rural income gap: does urban spatial form matter in China? SAGE Open; 2021.

  64. Maslow AH. A theory of human motivation. Psychol Rev. 1943;50(4):370–96.

    Article  Google Scholar 

  65. McAuliffe E, Galligan M, Revill P, Kamwendo F, Sidat M, Masanja H, et al. Factors influencing job preferences of health workers providing obstetric care: results from discrete choice experiments in Malawi, Mozambique and Tanzania. Global Health. 2016;12(1):86.

    Article  Google Scholar 

  66. Gyarteng-Mensah H, Owusu-Manu DG, Edwards D, Baidoo I, El-Gohary H. An assessment of students’ job preference using a discrete choice experiment: a postgraduate case study. J Humanit Appl Soc Sci. 2022;4(3):159–78.

    Google Scholar 

  67. Ding Z, Xia Y, Zhang L. Intergenerational support for the younger ageing parents in only child families in urban China: a comparison with multiple children families. Popul Res. 2019;43(2):87–99. (in Chinese).

    Google Scholar 

  68. Song J, Huang F. Intergenerational relationship between adult only child and their parents. Popul Res. 2011;35(3):3–16. (in Chinese).

    Google Scholar 

Download references

Acknowledgements

The authors would like to thank all the students who voluntarily participated in this study.

Funding

This study was supported by the “14th Five-Year” Medical Education Science Research Project of China Medical University (YDJK2021062).

Author information

Authors and Affiliations

  1. Department of Social Medicine, School of Health Management, China Medical University, No. 77 Puhe Road, Shenyang North New Area, 110122, Shenyang, Liaoning, China

    Nansheng Luo, Ru Bai, Yu Sun, Xueying Li, Libing Liu & Li Liu

  2. Department of Clinical Epidemiology, Shengjing Hospital of China Medical University, No. 36 Sanhao Street, Heping District, 110004, Shenyang, Liaoning, China

    Xin Xu

Authors
  1. Nansheng Luo
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ru Bai
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Yu Sun
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Xueying Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Libing Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Xin Xu
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Li Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Nansheng Luo was responsible for data analysis and paper writing and editing. Ru Bai and Yu Sun were responsible for data cleaning, assisting in data analysis and interpretation of the data. Xueying Li and Libing Liu were responsible for data collection, the preliminary arrangement of data, and assisting in data interpretation. Xin Xu and Li Liu provide theoretical guidance and methodological guidance, and design of the study. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Xin Xu or Li Liu.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of China Medical University and all methods were carried out in accordance with relevant guidelines and regulations. Participants were well-informed about the purpose and protocol of the study, and informed consent was obtained from them. The information gathered will be used exclusively for research purposes.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1: Supplementary Table 1.

Results of mixed logit models and WTP (Single child). Supplementary Table 2. Result of mixed logit models and WTP (Place of origin). Supplementary Table 3. Result of mixed logit models and WTP (Annual family income (RMB))

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Luo, N., Bai, R., Sun, Y. et al. Job preferences of master of public health students in China: a discrete choice experiment. BMC Med Educ 24, 24 (2024). https://doi.org/10.1186/s12909-023-04993-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12909-023-04993-9

Keywords

BMC Medical Education

ISSN: 1472-6920

Contact us