Development of preservice teachers’ competence beliefs, career values, and perceptions of teaching as a profession: A longitudinal study

Abstract

Competence beliefs, career values, and perceptions regarding teaching as a profession have important implications for preservice teachers' professional development. Longitudinal research concerning the development of preservice teachers' competence beliefs, values, and perceptions of teaching remains scarce. We investigated a sample of 270 Chinese preservice teachers on three measurement occasions. The expectancy-value theory framework was employed as a theoretical framework. We first tested the longitudinal measurement invariance of the FIT-Choice scale and then examined the growth trajectories of preservice teachers’ competence beliefs, career values, and perceptions of teaching as a profession over time. The results showed that the FIT-Choice scale was characterized by longitudinal measurement invariance. The Chinese preservice teachers included in our sample exhibited an overall decline in the growth trajectories of values (e.g., social utility value) and perceptions of teaching (e.g., expertise, social status, salary, and career choice satisfaction) over time, with the exception of their perceptions of difficulty, which exhibited an overall increase over time. Their initial levels of all types of values and perception constructs exhibited individual variability. We did not find individual differences in the development of the constructs that we studied. Practical implications are discussed.

1. Introduction

Motivations play a crucial role in influencing students' career choices, school engagement, and achievement in the context of educational psychology [1]. A number of recent studies have shown that preservice teachers’ motivations for choosing to become teachers and their perceptions concerning teaching as a profession are important factors related to their career choice satisfaction and future professional engagement [[2], [3], [4], [5]]. However, little is known regarding the development of these factors over time throughout the course of undergraduate studies in the context of teacher education.

Most studies concerning preservice teachers' motivations and perceptions regarding the teaching profession have been cross-sectional. Few longitudinal studies have been conducted. In the present study, we focus on the developmental trajectories of preservice teachers' motivations and perceptions during their first and second years of college (i.e., whether they increase, decrease, or remain stable). Given that preservice teachers’ motivations and perceptions have important educational and professional implications, the findings of this study shed light on the processes underlying teacher education programs for quality teachers.

1.1. The expectancy-value theory framework

Expectancy-value theory is one of the most prominent theories of motivation used in educational psychology, and it has been employed widely to understand academic behaviors, achievement, and achievement-related choices in many different fields [6,7]. This theory posits that task values and expectancies for success (or competence beliefs) are the most important motivating factors that predict academic choices and achievement. Task values have four subcomponents: intrinsic value (i.e., the enjoyment or pleasure one would obtain from the performance of a task), utility value (i.e., how useful a task would be to an individual), attainment value (i.e., whether one would consider a task to be personally meaningful and important), and cost (i.e., the sacrifices or effort that one would have to make or expend). Expectancies for success include ability beliefs (i.e., individual perceptions of one's competence at a given task) and expectations regarding success (i.e., how well one can perform with respect to upcoming tasks). In general, task values are reliable predictors of academic or career choices, and competence beliefs are frequently associated with engagement and performance [8,9].

The core of the Factors Influencing Teaching Choice (FIT-Choice) framework—“self, value, and task perceptions”—was developed based on the key constructs of expectancy-value theory [10]. These authors also added antecedent (i.e., social influence) and consequence (i.e., career choice satisfaction) variables to adapt the FIT-Choice framework more specifically to the context of teacher education. The original names of the constructs used in expectancy-value theory were relabeled, and new items were developed accordingly to clarify interpretations regarding an individual's choice of teaching as a profession. For example, what was originally known as attainment value was renamed personal utility value to measure the personal aspects of preservice teachers' regarding the values that are relevant to a teaching career. Utility value was renamed social utility value to assess the extent to which a teaching career is regarded as socially useful. The task perceptions construct was divided into task demand and task return subconstructs to measure preservice teachers' perceptions of teaching.

1.2. The development of competence beliefs, career values, and perceptions over time

For children and adolescents (i.e., elementary school, middle school, or high school students), it is already known that their competence beliefs and task values exhibit an overall decline in many ways during their school years [11,12]. However, studies conducted in the context of higher education have revealed a different picture—university students' competence beliefs and task values can increase, decrease, or remain stable. A recent review found that university students' competence beliefs are higher under certain conditions than others and can be improved [13]. University students' competence beliefs in quantitative and humanities subjects increased during their first year of college [14]. On the other hand, Kosovich, Flake, and Hulleman found that both university students’ competence beliefs and their utility values declined over the short term and that different individuals exhibited variability in terms of growth [15]. Robinson et al. also reported that among engineering students during their first 2 years of college, competence beliefs and three types of task values decreased, whereas perceived cost (opportunity, effort, psychological) increased [16].

In the context of teacher education, studies often show mixed findings of change and stability in the relatively short time frame, while another story of change in the long run. On one hand, Taimalu, Luik, and Täht found that competence beliefs, social influence motivations, and perceptions of expertise increased, while most other factors remained stable, among preservice teachers tracked from the beginning to the end of first year in university [17]; a qualitative study in Hong Kong also showed that while some preservice teachers developed their motivations towards teaching in the course, some others showed little change [18]. One the other hand, Bargmann, Thiele, and Kauffeld found that preservice teachers’ competence beliefs increased, and their interest value decreased after the first year in higher education [19]. Changes in motivation could happen to some extent over a semester, particular as a result of teaching practicum [20,21]. Rots, Kelchtermans, and Aelterman reported changes during a three-year teacher education program were also reported in a qualitative study conducted in Belgium [22].

Compared to practicing teachers, preservice teachers perceive less overall stress and lower confidence with respect to managing classroom behavior [23], and they tend to have naïve and idealistic perceptions of teaching [24]. Therefore, it is possible that their motivations for and perceptions of teaching are undergoing a developmental process in which they are formed and shaped through ongoing experiences and interactions with current circumstances.

1.3. The present study

In the present research, we employ a longitudinal design to address the following two research questions. First, does the FIT-Choice model [10] exhibit longitudinal measurement invariance over time? Second, how do preservice teachers' competence beliefs, career values, and perceptions regarding teaching as a profession change over time (i.e., what are their developmental trajectories)? In addition, are there individual differences in their initial levels of and growth rates in these studied constructs? To answer these questions, we investigated a sample of Chinese preservice teachers in their first or second years of college and measured their competence beliefs, career values, and perceptions on three occasions. We focused on first- and second-year preservice teachers because new college students may choose a program to study with no or little career-related experience, and they frequently have unrealistic perceptions or expectations during their transitions to university [25,26]. We also focused on competence beliefs, career values, and perceptions of teaching as a profession because these factors play important roles in preservice teachers' career decidedness and professional engagement [19]. The growth trends exhibited by competence beliefs, career values, and perceptions of teaching as a profession may be indicative of their academic performance or professional development. In light of the declining growth patterns revealed by the expectancy-value literature, we anticipated that preservice teachers’ competence beliefs, career values, and perceptions of teaching would exhibit an overall decrease over time in this study.

2. Methods

2.1. Context

The teacher education process in China is similar to other OECD member countries [27]. Faculties of education at normal universities offer three-year associate degree programs or four-year bachelor's degree programs to students who want to teach in primary or secondary schools [28]. In their first two college years, preservice teachers take a series of foundational courses (e.g., political theories, language [Chinese or English], and advanced mathematics); in the third or fourth year, they are required to take pedagogical courses (e.g., teaching methods and principles, educational psychology, and research methods in education). They are sent to primary or second schools for teaching practicum, usually in the third spring semester. Although some preservice teachers do take part-time jobs tutoring, most first- or second-year preservice teachers do not have a real teaching experience in classrooms.

2.2. Sample

Data were collected from a sample of preservice teachers who were studying in a four-year degree program at a typical normal university in Shandong, China, based on a convenience sampling approach. The survey was conducted in November 2020 (T1) (i.e., during the fall semester of their first academic year in college), May 2021 (T2) (i.e., during the spring semester of the same academic year), and October 2021 (T3) (i.e., during the fall semester of the following academic year). At T1, the participants were 270 first-year preservice teachers (including 40 [14.8%] males and 230 [85.2%] females; mean age = 18.31, SD = 0.73 years). At T2 (N = 250), 20 preservice teachers did not participate. At T3 (N = 144), 126 preservice teachers did not participate. The attrition rate was 7.4% for the second assessment wave and 46.7% for the third wave. The longitudinal attrition at the final time point might be due in part to the transfer of the project investigator to another university at that time. We also conducted longitudinal attrition analyses via logistic regression analysis to examine the extent to which attrition at T3 was related to the variables studied at T1 and T2. The results showed that males (B = −1.37, p = 0.002, OR = 0.25) were more likely to withdraw from the study. Special efforts to retain male preservice teachers should be implemented in future longitudinal studies.

2.3. Measures

The FIT-Choice scale [10] was used to gauge preservice teachers’ motivational orientations toward and perceptions regarding teaching. Competence beliefs were assessed with three items (e.g., “I have good teaching skills”). Career values included three subscales: intrinsic career value was measured with 3 items (e.g., “I like teaching”); personal utility value was measured by job security (3 items, e.g., “Teaching will be a secure job”), time for family (5 items, e.g., “As a teacher I will have lengthy holidays”), and job transferability (3 items, e.g., “A teaching job will allow me to choose where I wish to live”); social utility value was measured by social contributions (e.g., “Teachers make a worthwhile social contribution”), enhancing social equity (e.g., “Teaching will allow me to benefit the socially disadvantaged”), working with children/adolescents (e.g., “I like working with children/adolescents”), and shaping the future of children (e.g., “Teaching will allow me to influence the next generation”), with 12 items in total. Perceptions of teaching were assessed by expertise (3 items, e.g., “Do you think teachers need highly specialized knowledge?”), difficulty (3 items, e.g., “Do you think teachers have a heavy workload?”), social status (6 items, e.g., “Do you believe teaching is a well-respected career?”), and salary (2 items, e.g., “Do you think teaching is well paid?”). A total of 46 items of the FIT-Choice scale were used in the survey. The scale responses ranged from 1 (not at all) to 7 (extremely). For details, please refer to the full questionnaire online.

The FIT-Choice scale was validated in the Chinese context by a previous study [2] (An et al., 2020). In the present study, the results of confirmatory factor analyses exhibited an overall acceptable model–data fit across the three time points (please refer to Table S1 in the supplemental file). All subscales exhibited good reliability, with Cronbach's alpha ranging from 0.78 to 0.97 across the three time points (Table 1).

Table 1.

Descriptive statistics and Cronbach's alpha values among the variables (n = 144–270).

	Variables	M	SD	S	K	α
Competence Beliefs	CB T1	5.32	1.35	−0.90	0.77	0.87
	CB T2	5.14	1.33	−0.67	0.62	0.89
	CB T3	5.32	1.35	−0.90	0.77	0.88
Career Values	ICV T1	5.42	1.60	−0.99	0.34	0.93
	PUV T1	5.06	1.34	−0.79	0.55	0.88
	SUV T1	5.68	1.35	−1.38	1.99	0.94
	ICV T2	5.15	1.48	−0.62	0.09	0.93
	PUV T2	5.22	1.21	−0.76	0.83	0.87
	SUV T2	5.47	1.31	−1.01	1.19	0.95
	ICV T3	5.42	1.49	−0.97	0.72	0.93
	PUV T3	5.16	1.23	−0.88	1.53	0.89
	SUV T3	5.58	1.33	−1.27	1.99	0.97
Perceptions	Expertise T1	6.02	1.34	−1.84	3.69	0.94
	Difficulty T1	4.92	1.32	−0.61	0.46	0.78
	Social Status T1	5.27	1.22	−0.83	1.38	0.86
	Salary T1	4.95	1.44	−0.49	0.08	0.89
	Expertise T2	5.76	1.31	−1.24	1.80	0.94
	Difficulty T2	5.06	1.21	−0.38	0.37	0.84
	Social Status T2	5.16	1.15	−0.47	0.69	0.87
	Salary T2	4.85	1.31	−0.14	−0.26	0.89
	Expertise T3	5.69	1.45	−1.45	2.31	0.97
	Difficulty T3	5.17	1.25	−0.76	1.13	0.86
	Social Status T3	5.13	1.22	−0.61	0.87	0.87
	Salary T3	4.63	1.51	−0.36	0.06	0.91
Career Choice Satisfaction	CCS T1	5.71	1.43	−1.23	1.08	0.91
	CCS T2	5.53	1.41	−0.94	0.73	0.95
	CCS T3	5.54	1.42	−1.16	1.44	0.97

Note. CB = competence beliefs; CCS = career choice satisfaction; S = skewness; K = kurtosis; ICV = intrinsic career value; PUV = personal utility value; SUV = social utility value.

2.4. Statistical analyses

The data were analyzed using SPSS version 23.0 and MPlus version 7 software for Windows. Little's test indicating that the missing data appeared to be completely random (p > 0.05). Therefore, we used the maximum likelihood estimator to address the issue of missing data by reference to all the available information [29]. The indices used to estimate the goodness of model–data fit and their thresholds were the chi-square goodness of fit value (p > 0.05), the comparison of fit index (CFI ≥0.90), the Tucker‒Lewis index (TLI ≥ 0.90), the standardized root mean residual (SRMR ≤ 0.10) and the root mean square error of approximation (RMSEA ≤ 0.08) [30].

Longitudinal measurement invariance (LMI) is a precondition for valid comparisons at the mean level. Configural, metric, and scalar invariance are three levels of LMI that have been routinely tested in the literature [31]. We tested LMI for each construct during the first stage of the data analyses in a stepwise manner [32]. We first tested the configural invariance to determine whether the factor structure was similar across different measurement occasions. No constraints were imposed during this step. In the subsequent step, metric invariance was tested by constraining factor loadings to be equal across different measurement occasions. Finally, we tested scalar invariance by additionally requiring item intercepts to be equal across different measurement occasions. When full LMI is unlikely to be achieved, some parameters can be allowed to differ instead of requiring all parameters to be equal to establish partial LMI [33]. LMI is indicated when the difference in the comparison of fit index (CFI) between the two nested models is equal to or less than 0.01 [34].

During the second stage of data analyses, latent growth curve models (LGCMs) were implemented for each construct separately to model appropriate development trajectories for each construct and to determine whether our participants exhibited individual differences in their overall levels of each construct as well as their growth rates. We first conducted a linear LGCM in which the factor loadings of the latent intercept were set at 1 and the factor loadings of the slope were fixed at 0.00, 1.00, and 2.00 for the three time points in order. If a linear LGCM was unable to fit the data, we considered an unspecified LGCM by setting the loadings of the slope for T1 and T3 at 0 and 1, respectively, while the loading for T2 was estimated freely.

2.5. Procedure

This longitudinal study underwent an ethics review and was approved by the corresponding author's institutional review committee. The data collection procedure consisted of three assessment waves. At each assessment wave, we first explained the objectives of the study and asked if preservice teachers would like to participate. After agreeing to participate, questionnaires were sent to them via a popular online survey platform (WJX, http://www.wjx.cn/) in class and took between 10 and 20 min to complete. Consent forms were also obtained as part of the data collection. Voluntary participation, confidentiality, and the freedom to withdraw at any point in time without explanation were guaranteed throughout the process of data collection. The data collection procedure was consistent across all three time points of the study.

3. Results

We tested measurement invariance for each construct across the three assessment waves. Taking the intrinsic career value construct as an example, the results first showed that the configural invariance model exhibited a reasonable fit (χ² (24) = 61.81, CFI = 0.98, TLI = 0.97, SRMR = 0.04, and RMSEA = 0.07); we further found metric invariance when constraining the factor loadings to be invariant across the three time points (ΔCFI = 0.00), and an additional constraint of the item intercepts to be invariant showed a small and acceptable decrease in the goodness-of-fit indices (ΔCFI = 0.01). Thus, as a whole, this evidence indicated strong longitudinal measurement invariance across the three waves of assessment. Fit statistics for models representing different degrees of invariance can be found in Table S2 in the online supplemental file.

In terms of growth analysis, the model coefficients of the LGCMs are presented in Table S3 in the online supplemental materials. The linear growth curve models fit the data well for all the constructs studied; therefore, we did not further test the unspecified LGCMs (i.e., the nonlinear models). Our participants differed in terms of their initial levels of every construct included in the study, as indicated by the significant variances of the intercepts. On the other hand, these participants did not differ in terms of the growth rates of every construct over time, as indicated by the nonsignificant variances of the slopes. In addition, the constructs exhibited decreases across these time points, as indicated by the significant negative slopes (social utility value: b = −0.11, p = 0.046; expertise: b = −0.2, p = 0.003; salary: b = −0.17, p = 0.008; career choice satisfaction: b = −0.13, p = 0.017), with the exception of difficulty, which exhibited an increase over time and thus a significant positive slope (b = 0.12, p = 0.046). Please see Table 2 for detailed growth parameters. We also present the developmental trajectories of variables with significant slope variances based on factor means in Fig. 1. Although mean scores of the variables with significant slopes for female participants were almost always higher than male participants across the three time points (excepting for difficulty at T1), independent samples t-tests results did not indicate any significant differences in these variables by gender (please refer to the online supplemental file).

Table 2.

Model parameters and fit indices of the latent growth curve models for values and perceptions.

Constructs	Means						Variance						Correlations			Fit Indices
	Intercept	SE	P	Slope	SE	P	Intercept	SE	P	Slope	SE	P	I with S	SE	P	CFI	χ2 (df)	P
Competence
Beliefs	5.29	0.08	<0.001	−0.08	0.05	0.120	0.85	0.25	0.001	−0.03	0.15	0.863	−0.05	0.17	0.798	0.99	1.69 (1)	0.194
Career Values
ICV	5.34	0.09	<0.001	−0.08	0.06	0.145	1.20	0.33	<0.001	−0.20	0.20	0.303	0.04	0.23	0.856	0.96	5.85 (1)	0.016
PUV	5.11	0.08	<0.001	0.05	0.05	0.378	0.64	0.21	0.002	−0.03	0.13	0.827	0.06	0.15	0.698	0.98	2.94 (1)	0.086
SUV	5.64	0.08	<0.001	−0.11	0.06	0.046	0.60	0.24	0.013	−0.08	0.15	0.622	0.08	0.17	0.656	0.99	1.51 (1)	0.219
Perceptions
Expertise	6.01	0.08	<0.001	−0.20	0.07	0.003	0.82	0.25	0.001	0.29	0.17	0.076	−0.28	0.19	0.128	0.99	1.19 (1)	0.274
Difficulty	4.93	0.08	<0.001	0.12	0.06	0.046	0.50	0.23	0.028	0.17	0.15	0.277	−0.15	0.23	0.03	1.00	0.07 (1)	0.792
SS	5.27	0.07	<0.001	−0.10	0.05	0.074	0.57	0.18	0.002	0.17	0.12	0.158	−0.08	0.13	0.563	1.00	0.20 (1)	0.655
Salary	4.98	0.08	<0.001	−0.17	0.06	0.008	0.69	0.24	0.003	0.15	0.16	0.321	−0.02	0.17	0.898	1.00	0.96 (1)	0.328
CCS	5.69	0.08	<0.001	−0.13	0.06	0.017	0.82	0.28	0.003	−0.14	0.16	0.405	0.08	0.20	0.684	1.00	0.26 (1)	0.607

Note. ICV = intrinsic career value; PUV = personal utility value; SUV = social utility value; SS = social status; CCS = career choice satisfaction.

*p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 1.

Developmental Trajectories of the Variables with Significant Slopes Based on Factor Means. Note. SUV = social utility value; Satisfaction = career choice satisfaction.

4. Discussion

By reference to a longitudinal dataset collected from a sample of Chinese preservice teachers across three measurement occasions, the present study aimed to investigate the longitudinal measurement invariance of the FIT-Choice model [10] and to model the developmental trajectories of participants’ competence beliefs, career values, and perceptions of a teaching career. This study provided important insights into the development of the factors motivating preservice teachers over time and yielded important practical implications.

4.1. Longitudinal measurement invariance of the FIT-Choice framework

The first major finding of our study supported the longitudinal measurement invariance of the FIT-Choice framework—the model exhibited consistent psychometric properties across the three measurement occasions. Configural invariance was first established, indicating that the preservice teachers included in the study exhibited similar understanding of competence beliefs, values, and perceptions over time. Metric invariance was further verified, i.e., different aspects of each of these constructs as measured by the items within that construct had the same relative importance over time. Finally, the FIT-Choice model also exhibited scalar invariance, indicating that when preservice teachers included in the study actually exhibited similar levels of, e.g., expectancy as measured across the three time points, the ratings collected for these items were also at similar levels. In summary, these findings suggest that this scale can be used to measure or compare longitudinal mean changes.

4.2. The development of preservice teachers’ competence beliefs, career values, and perceptions over time

Participants in the study maintained the same levels of competence beliefs over time, which is in line with several studies [[35], [36], [37]]. For example, preservice teachers' beliefs regarding EFL teaching ability remained stable throughout the first term of the teacher training program [35,37]. In the studied context, first- and second-year preservice teachers mainly take theoretical courses and frequently lack opportunities to observe and teach in real classrooms, and it is usually difficult for them to develop awareness of competence beliefs [36]. However, several studies [17,19] showed that preservice teachers' competence beliefs increased in some circumstances, at the end of their first year in university. Although teaching practicums often serve as a significant source of preservice teachers' competence beliefs [38], an initial sense of confidence in their ability to teach during their early years in university is also important because preservice teachers' competence beliefs mediate the relationship between career adaptability and career optimism [39]. Educational strategies to promote preservice teachers’ competence beliefs in the context of theoretical courses are needed.

In terms of career values, participants' social utility value declined, while the remaining types of career values remained stable. This negative change in social utility value implies that a teaching career became less socially useful to participants. Their altruistic reasons for choosing a teaching career weakened during a teacher preparation program. In addition, Bargmann et al. found that preservice teachers’ interest value decreased during their first year in university (these authors did not test other types of career values) [19]. Overall, this evidence sends a message that preservice teachers come to value a teaching career less in certain respects, at least during their early years in university. Teacher educators are also advised to apply expectancy-value theory to daily teaching and to design a sustained motivating learning environment for preservice teachers.

Although preservice teachers' motivations generally remained stable in this study, each and every type of perceptions regarding teaching became more negative. That is, our respondents perceived teaching more as a difficult, low-paid job while requires a high level of expertise over time. This may imply that teaching as a profession became less attractive to them. Preservice teachers at this stage neither assumed the role of a teacher nor had real contact with students in schools; thus, they could have naïve or unrealistic perceptions regarding teaching prior to participating in teaching practicums [24]. After accepting good students into a teacher preparation program, additional challenges continue to ahead, and efforts to cope with such challenges remain difficult. It would be encouraging to see preservice teachers’ career perceptions developing positively during their senior years in university, especially when they participate in teaching practicums.

Watt and Richardson found that less engaged beginning teachers became less satisfied with their choice of teaching as a career throughout their experience with teacher education [40]. Our study additionally revealed that preservice teachers' career choice satisfaction started to decline during their early years in university. Because career choice satisfaction has important implications for individuals' professional engagement and well-being [[41], [42], [43]], a pattern of declining satisfaction is detrimental to both preservice and in-service teachers in many ways. Why did people's levels of satisfaction show an overall decrease immediately following their enrollment in a teacher education program in this study? Why would some students' levels of satisfaction increase while those of others decrease? Less is known regarding the factors that may influence longitudinal changes in preservice teachers' career choice satisfaction [40], especially during their early years of teacher education.

Previous work in teacher education has ignored the difference between linear and nonlinear patterns of development in preservice teachers' motivations and values. Although our findings revealed linear growth patterns, nonlinear growth is also possible if the time frame is extended to four years. In the Chinese teacher education system, students usually participate in teaching practicums during the spring semester of their third academic year. Prior to that time, preservice teachers take theoretical courses, and most of these individuals do not have the opportunity to teach in classrooms (although some preservice teachers do take part-time jobs tutoring). Because the educational experience of a practicum is central to preservice teachers' professional development [[44], [45], [46]], it is possible that some types of motivation remain stable during preservice teachers' first two years in university (as this study revealed) but ultimately exhibit an overall increase following the teaching practicum. For example, preservice teachers' self-efficacy changes on a weekly basis and increases during the teaching practicum [47]. It is also possible that some types of perceptions of teaching first decrease during preservice teachers’ first two years of study and subsequently decrease even further following the teaching practicum, given that preservice teachers must face the actual job demands and tensions associated with meeting student needs as well as the expectations of schools and parents during the teaching practicum [40].

4.3. Practical implications

Preservice teachers’ competence beliefs, career values, and perceptions of teaching as a career are important predictors of their career decidedness, intention to drop out, and career choice satisfaction [2,19]. Therefore, their change is a matter of concern to teacher educators and career counsellors.

The decline of some key constructs revealed by this study conveys important practical implications that future work is clearly needed. As the first step, institutions should track preservice teachers throughout their teacher education program to evaluate the change in competence beliefs, career values, and perceptions of teaching. Support service (e.g., counselling interviews) could be offered for those students who report negative changes. Second, efforts to promote or improve the perceptions of teaching among first- and second-year preservice teachers are especially necessary, as the findings of this study indicated that most of the decrease in these constructs occurred in perceptions. More information on real teaching practice could be provided to prospective high school students who want to choose a teacher education program, so that they could have more realistic perceptions at the beginning of teacher education. Orientation phases can also be helpful after students are admitted to universities. At the same time, teacher educators should consider that younger preservice teachers are less familiar with real teaching practice [24], and provide them with field experience through various activities like workshops or educational dramas rather than solely relying on teaching practicum, which usually has been performed during the third spring semester in the studied context. Finally, further exploration of the antecedents of these constructs are also necessary because the enhancement of these antecedents may help preservice teachers maintain or improve their levels of competence beliefs, career values, and perceptions.

4.4. Limitations and directions for future research

Although the present study provides new insights into the development of preservice teachers' competence beliefs, career values, and perceptions of teaching as a profession over time, several limitations of this study should be acknowledged and could be addressed by future studies. First, the study revealed a general decline in Chinese preservice teachers' perceptions of teaching over time. A further question pertains to whether and how teacher educators can help teachers maintain or improve their perceptions of teaching over time. Intervention programs based on empirical findings concerning the predictors or antecedents of perceptions of teaching would be more effective in this context. Second, the assessments referenced by the study were conducted every semester, and the time interval between adjacent time points was half a year, which was a relatively short duration. A longer time frame (e.g., from admission to graduation) might facilitate additional understanding of teachers' development of their values and perceptions throughout the entirety of their undergraduate teacher education. It is possible that preservice teachers begin their college studies by experiencing a decline in their values and perceptions during the first or second year in college but that those values and perceptions subsequently increase during the third or final year of college alongside the teachers' deepening understanding of teaching as a profession. In addition, although this study did not find variability in terms of the growth rates of the constructs we examined during this shorter time period, it is possible that teachers’ growth rates vary over longer periods of time. Finally, these findings were based on a relatively small sample of Chinese preservice teachers, which was not nationally representative. It would be valuable to study a larger sample and include school-level variables in multi-level studies. Generalization of such findings to other contexts should be done with caution. Further studies focusing on other cultural contexts are also necessary.

5. Conclusions

The findings of this study led to at least four conclusions. First, the FIT-Choice theoretical model [10] demonstrated longitudinal measurement invariance over time, thus suggesting that it can be used for longitudinal comparisons. Second, the preservice teachers whom we followed from their first to their second year of college exhibited an overall decline in the growth trajectories of their values (e.g., social utility value) and perceptions (e.g., expertise, social status, salary, and career choice satisfaction) of teaching over time, with the exception of their perceptions of difficulty, which showed an overall increase over time. Third, we found individual variability in the initial levels of all types of competence beliefs, career values, and perception constructs. Fourth, we found no individual differences in the development of the constructs that we studied.

Author contribution statement

Min An: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Xiaofei Yu: Conceived and designed the experiments; Analyzed and interpreted the data; Wrote the paper.

Xiaofeng Zhao: Performed the experiments; Contributed reagents, materials, analysis tools or data.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability statement

Data will be made available on request.

Declaration of interest's statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: