Factors Associated with the Development of Secondary School Students’ Interest towards STEM Studies

Abstract

The aim of this study is to investigate the predictive factors that predispose secondary school students’ interest in studying STEM (Science, Technology, Engineering, and Mathematics) fields in higher education. For this purpose, an already existing questionnaire was used and modified properly, according to the Greek educational system. The survey was attended by 301 secondary school students, who study in Piraeus, one of the largest cities in Greece. Research findings indicated that the principles of Social Cognitive Career Theory (SCCT) are well supported. It is worth mentioning that this is the first time that such a number of variables had been examined, in order to support the SCCT. In particular, very few studies exist in literature—to the best of our knowledge—investigating the effect of more than four factors influencing students’ interest towards STEM higher studies. Learning experiences, students’ exposure to STEM activities within the school environment and outside of it (OR = 0.071, p = 0.002), as well as their involvement with high difficulty STEM courses (OR = 0.203, p = 0.038), appear to be positively correlated with the development of interest towards studies in the STEM fields. In addition, students from low-income families are more likely to follow STEM studies (OR = 0.198, p = 0.034). On the contrary, it has been revealed that parental educational background only supports the student’s decision to continue studies after high school, without specifying the educational field in higher education (OR = 0.769, p = 0.703; father’s educational level, OR = 0.698, p = 0.552; mother’s educational level). Data revealed that outcome expectations and self-efficacy (OR = 14.366, p = 0.005) are positively related to the procedure of students’ interest development to pursue STEM fields in higher education, while gender seems to be a non-regulatory factor (OR = 0.886, p = 0.831).

1. Introduction

Social Cognitive Career Theory (SCCT) is an important conceptual and theoretical scheme that supports the holistic study and exploration of those factors that act in the cognitive process of making educational and professional decisions. The theory was founded and formulated by Lent, Brown, and Hackett [1,2,3]. According to researchers, the contribution of individual and intra-patient situations to the successful outcome of requirements linked to external situations should be seriously considered and inextricably linked to the expectation of a result or “self-efficacy”. The term “self-efficacy” was first introduced by A. Bandura, who mentioned it in many studies, including that of the Social Learning Theory [4,5,6,7,8,9,10]. However, many investigations about “self-efficacy” have also been performed by other researchers [11,12]. The meaning of “self-efficacy” or “self-efficacy expectation” can be also looked up in Social Cognitive Theory of Behavior [13,14], which assesses the person’s ability to define himself (human agency). Bandura’s theory emphasizes indirect learning (vicarious learning), which is the learning that occurs by observing others. Simultaneously, that type of learning can lead to the development of the socio-cognitive mechanisms of an individual, in order to regulate their behavior for future success or performance.

The main point that differentiates SCCT from other models of professional development is the emphasis on the existence and functioning of social-cognitive mechanisms as determinants of human behavior. Based on other professional development models, a person’s behavior is considered to be a derivative of the interaction between the person and the environment. In SCCT, however, human behavior is also considered as a co-modifier of the environment. Hence, SCCT adopts the triadic reciprocal model of causality of A. Bandura [15], whereby the cognitive and environmental factors, as well as the manifest behavior of a person, constitute mechanisms that interact dynamically with each other by influencing each other.

Τhe theory in the first part focuses on the cognitive factors that are supposed to shape both the professional interests and choices and the consequent performance. Nevertheless, it is reasonable that human actions do not take place in an environmental vacuum. For this reason, in the second part of the theory, the rapporteurs deal with the role of individual and environmental factors. Therefore, for the SCCT’s rapporteurs, an individual’s professional development “is a picture of theatrical drama that” rises “on the social stage” [1]. In this context, personal, environmental, and learning factors are highlighted, acting as crucial contributions to the individual’s career path. Through these factors, the central cognitive mechanisms (such as self-efficacy perceptions and outcome expectations) as well as the individual’s consequent professional development are being shaped. [2]. It is noteworthy that SCCT is also influenced by other academic and occupational behavior models [16,17,18], as well as by Vroom’s Expectation Theory [19] and Locke [20] and Locke and Latham’s Goals Setting Theory [21]. In addition, it is inspired by theories that focus on the influence either of environmental factors [22,23] or genetic factors on behavior and personality [24,25]. Researchers’ findings [11,26,27,28,29] have shown that the relationship between self-perceptions of mathematics, technology, and science with gender, educational, and professional aspirations and choices has been investigated in students in the aforementioned STEM fields. SCCT has been applied to a wide range of studies related to STEM education at school [25,30,31]. Moreover, SCCT has been applied to students, in order to predict their interest both in STEM education and STEM career choice [25,32].

In Greece, Secondary Education is provided in two cycles: Compulsory Secondary Education offered by “Gymnasium” and Post-secondary Secondary Education provided by the “Unified High Schools” and “Technical Vocational” Schools. It should be noticed that “Gymnasium” consists of three grades with total duration of three years, corresponding to the 7th, 8th, and 9th grade. The curriculum consists of General Education courses, which are the same for all the students.

It should also be noticed that the “Unified High School” consists of three grades with a total duration of three years. In the 1st grade of the “Unified High School”, which corresponds to the 10th grade of the international educational system, the curriculum consists of General Education courses, which are the same for all the students. In the 2nd grade of the “Unified High School” (11th grade), students have to choose one of the two following directions: “Theoretical” or “Positive”. During this grade, students are taught both general and directional education courses. “Positive” direction courses include Mathematics and Physics. Finally, in the 3rd grade (12th grade) of the “Unified High School”, students have to choose one of the following directions: Humanitarian Sciences, Positive Sciences, and Economics. The courses of “Positive” direction are Mathematics, Physics, Chemistry, and Biology. Students of the Economic Cycle are also taught Economics and Informatics.

In our research model the main six variables studied, according to SCCT, are gender, family background, learning experiences, self-efficacy, outcome expectations, and social support. Our findings confirm that the principles of SCCT are well supported. Limited studies exist in literature—to the best of our knowledge— which study the effect of more than four factors influencing students’ interest towards STEM higher studies.

2. SCCT Predicting Factors of Students’ Interest in Higher STEM Education

The main six variables, which predict students’ interest in higher STEM education are represented below. In this study, it is clarified that higher education is considered to be the attendance at an academic institution, such as a University or a College, leading to the acquisition of a degree or diploma. Additionally, fields such as Medicine, Mathematics, Physics, Chemistry, Biology, Statistics, Computer Science, Agricultural, Engineering Sciences, and Architecture were defined as STEM fields.

Gender: Several surveys have shown a high level of self-efficacy for men both for traditional males’ and females’ work. For female students, however, high levels of self-efficacy have been noted solely for female occupations [12,33]. Other researchers have demonstrated that different gender socialization contributes to the creation of differential perceptions of self-efficacy and, consequently, different perceptions for career choices [34,35,36]. Further studies have pointed out that the effect of gender on professional development stems from both personal physiological skills and the degree of acceptance by the socio-economic context [2]. Following the same theoretical basis, the concepts of self-efficacy in Science and Mathematics were explored. Within this framework, it was revealed an influence of gender, both on interests and the choices for studying in STEM fields and justified the under-representation of women in these specific cognitive fields [37,38]. In 2001, Bandura and his colleagues [14], have shown that self-efficacy perceptions differ in terms of gender. Boys have a high level of self-efficacy perceptions for male occupations in science and technology, while girls for women activities, like social and educational ones [39].

Rollins and Valdez’s recent surveys [40] on a group of adolescent pupils in terms of gender and socio-economic backgrounds, demonstrated higher levels of girls’ self-efficacy perceptions for the educational and professional decision-making process for “traditional” female professions (i.e., music and social sciences). Additionally, Williams and Subich [41] discovered that men reported greater access to information on realistic and research-based professional occupations, whereas women were mainly interested in Social Sciences careers, along with a great lack of interest in Mechanics [42,43].

Family background: Since the socio-economic background of the family belongs to the environmental factors and in particular to the contextual affordances, it precedes and formulates both educational and socio-cognitive mechanisms. Children of higher socio-economic background feel more comfortable with the selection process [44,45]. The higher the socio-economic level, the higher the aspirations of young people and their academic achievements [34,46,47,48,49]. Research data on Asian, African American, and Latin American adolescents have shown that the cultural atmosphere at home and the parental educational level play a key role in the professional preferences of young people [37,50]. Low socio-economic status, as a restrictive factor in the educational and professional decision-making process, is apparent, as denoted in several studies [1,51,52]. In particular, students from low socio-economic level were less interested in studying in STEM fields than their counterparts from higher socio-economic levels. However, students’ educational achievements reverse the aforementioned socio-economic difference [53].

Recent research has shown that working-class children pay attention to the external characteristics of study and work (professional opportunities, remuneration), while children of senior social classes have a more internalized perception of work [54]. Based on the fact that the former ones pay particular attention to the outcome expectations of studies and work [55], it is more possible to express greater interest in studying STEM fields than the latter ones. Similar results are also observed in the research of Jackson and Lichtenberg [56], as well as Lichtenberg and Jackson [57], on the basis that economically disadvantaged students are more confident in dealing with STEM fields. Finally, the study by Chachashvili-Bolotin et al. [58] indicates that the higher the students’ perception of their family financial background, the lower their interest in studying in STEM fields in higher education.

Learning experiences: A body of evidence has demonstrated that the exposure of students to STEM activities and programs concurrently into and outside the school environment strengthens their interest in STEM studies in higher education [59,60]. In addition, their engagement with STEM extra classroom activities helps them better understand these processes [61]. Indeed, the younger an individual is exposed to these actions, the more likely it is for them to make a STEM career choice [62,63]. Further studies have demonstrated similar findings [59]. Students’ educational achievements in mathematics seem to be related to their future engagement in studying STEM fields and pursuing a STEM career [25]. Although educational achievements in mathematics were considered to be one of the main predictors of STEM studies in higher education, further data have established that exposure to natural sciences is more decisive for future enrolment in STEM fields [25].

Self-efficacy: The term “self-efficacy” was introduced by A. Bandura in 1986 [15] and is considered as the subjective judgments and perceptions of an individual’s abilities, in order to successfully fulfill a task. Apart from personal beliefs, however, self-efficacy also includes all the actions need to be followed by the subject for a successful outcome (response outcome expectation), as well as the results of these actions. Self-efficacy plays a key role in the development of interests, ambitions, choices, and aspirations [64,65,66], which are fundamental factors for a certain behavior and performance [13,14,15,67,68].

Outcome expectations: Future positive profits of a particular choice will mobilize students towards a specific behavior, resulting in a certain outcome expectation. For instance, studying in STEM fields has been considered to be related to future professional opportunities, higher income, and well-respected social status [25].

Social support: Regarding the environmental mechanisms that function supportively, these are specific situations or environmental sources that facilitate the formulation of the individual and their effort to achieve their professional goals [1]. Examples of such mechanisms are financial support, the encouragement by important people in the family or friends, or interaction with people who can act as mentors in the life of the individual.

Based on the aforementioned theoretical framework, in this research the variables that were examined are gender corresponding to student’s personal data; the student’s socio-economic level; and the student’s social perceived support by direct and indirect social environment. These belong to the environmental factors and particularly contextual affordances; outcome expectations and self-efficacy (belonging to the cognitive mechanisms of the individual); as well as learning experiences (student’s exposure to STEM programs and/or activities in and out of school, student’s achievements, and/or performance at school when involved in STEM courses).

Research Question

Based on the above, the research design was realized in such way as to answer the following research question:

What are the SCCT-based factors that are able to predict secondary students’ interest in pursuing STEM fields in higher education?

3. Materials and Methods

We employed a self-designed questionnaire to measure the factors that are able to predict secondary students’ interest in pursuing STEM fields in higher education. The questionnaire included 11 questions regarding students’ attitudes towards higher education, 7 questions regarding students’ knowledge about higher education, 8 questions about personal information, and 5 questions about information regarding students’ background. The reliability coefficient Cronbach’s alpha was 0.598.

The sampling method applied in this study was convenience sampling. Prior to completing the survey, approval from the school principals was obtained. The potential participants were contacted during their regular school classes and were informed of the purpose of the study and the time demands. They were further informed that their responses would remain confidential and that their participation could be revoked at any time. Our survey was attended by 301 secondary school students, who study in eight Public Secondary Schools of Piraeus from 8th to 10th grade. 50.17% of the participants were boys, aged 14–16 years old. The study was conducted between September 2019 and November 2019, within the regular school schedule. The time needed to complete the survey was approximately 30–35 min.

Input Variables:

Table 1. Description of input variables in the study.

Input Variables
Dependent Variable	1. How important in your view is pursuing higher education?
	2. Disregarding your school grades, what is the area of study you would be interested to study in higher education?
Independent Variables	1 Gender
	2. Family background assessed via parents’ educational level and students’ perception of the family economic background
	3. Learning experiences:
	3.1 Student’s school performance in mathematics
	3.2 Student’s involvement in courses related to STEM education
	3.3 Student’s exposure to STEM activities and programs both in and outside of the school environment
	4. Self-efficacy
	5. Outcome expectations
	6. Student’s social support

presents the main variables of the study. The dependent and independent variables are described in detail below.

Dependent variable

The dependent variable is the interest of secondary school students in studying STEM fields in higher education. This variable was measured through the following two questions:

(1)

How important in your view is pursuing higher education? (see question 1 in Appendix A)

(2)

Disregarding your school grades, what is the area of study you would be interested to study in higher education? (see question 4a in Appendix A). Based on the answers to these two questions, the following three categories of students were distinguished:

1—I am not interested in studying at an academic institution, 2—I am interested in studying at an academic institution, but not related to a STEM field, and 3—I am interested in studying at an academic institution that is related to a STEM field.

Independent Variables

The independent variables considered were gender, family background (assessed through mother’s and father’s educational level and the students’ perception of the family economic background), learning experiences, self-efficacy, outcome expectations, and student’s social support. Specifically:

Student’s school performance in mathematics: a number of studies have shown that students’ high performance in mathematics is positively related to study in STEM fields [25]. In addition, the score in mathematics is of key importance for applicants, in order to be accepted into Technical Schools, Mathematics, and Natural Sciences. The questions are the following:

(1)

How much difficulty would you say that you have in attending mathematics in the school environment? (see question No. 23a of Appendix A)

(2)

What is your average score in Mathematics? (see question No. 24a of Appendix A). The variable was then divided into three categories, (a) low performance (below 59), (b) moderate performance (between 60–79), and (c) high performance (between 80–100).

Student’s involvement in courses related to STEM education such as Physics, Chemistry, Biology, and Computer Studies: the student’s involvement in the above-mentioned courses was measured by the following questions:

(1)

How much difficulty would you say that you have in attending Physics, Chemistry, Biology, and Computer Studies?

(2)

What is your average score in Physics, Chemistry, Biology, and Computer Studies? (see question No. 24b,c,d,e of Appendix A). The variable was then divided into three categories, (a) low performance (below 59), (b) moderate performance (between 60–79), and (c) high performance (between 80–100).

Student’s exposure to STEM activities and programs both in and out of the school environment: Students who have visited STEM academic institutions, attended STEM programs and/or clubs, or have taken private courses in STEM fields, were categorized into one analysis group, while those who were not exposed were categorized into another one. The questions related to this variable included:

(1)

Have you ever visited an Academic Institution during your school visits? (see question No. 11 of Appendix A). If the answer to this question is positive the students are required to clarify the institution with the following question:

(2)

Which one/ones? (see question No. 12 of Appendix A). The student is invited to select the institution they had visited from a list of academic institutions. Then, the selection was categorized into STEM or non-STEM field. By definition, institutions like the National Technical University of Athens, the Department of Positive Studies of the University of Athens (UoA), the Department of Informatics of UoA, the Medical and Pharmacy School of UoA, the Agricultural University of Athens, and the University of Piraeus were considered to be STEM institutions (answers 1-2-3-4-5-9 of question No. 12 of Appendix A) The other alternatives (answers 6-7-8-10 of question 12 of Appendix A) were considered as non-STEM institutions. Then, the variable was split into three categories: (a) they have not visited any higher academic institution, (b) they have not visited a STEM higher academic institution, and (c) they have visited a STEM higher academic institution.

(3)

In recent years, a number of activities, which take place in the school environment, are offered to young people and are supportive (encouraging) for them to pursue higher education. Please mark the activities you take part in from the list below. You can choose more than one activity. (see question No. 13 of Appendix A).

The proposed clubs of Physics, Chemistry, Mathematics, Robotics, Computer Science and Computer Engineering, and Biotechnology-Biochemistry (answers 1-2-3-4-5-6 of question No. 14 of Appendix A) were considered as activities that expose a young person to STEM fields. The other clubs were considered as non-STEM (answers 7-8-9-10-11-12-13 of question No. 13 of Appendix A). The order of preference of the above activities is to enhance the choice of students towards a particular academic direction and in a particular field of study. Based on the answers given to these questions, a three-tier variable was developed: (a) STEM activities, (b) non-STEM activities, and (c) no activities (answer 14 of question No. 14 of Appendix A).

(4)

Do you take any private lessons in the following subjects? (see question No. 25 of Appendix A). The students are required to define how many hours they take private lessons in Math, Computers, and Natural Sciences on a weekly or monthly basis. The answers of each subject were then divided into three categories: 1—Yes, on a weekly basis, 2—Yes, but less than once a week, 3—No, I am not taking any private lessons in these subjects.

Self-efficacy: was measured by the following question: “If you were going to follow studies in higher education in a field that you were not be interested in, do you believe that these studies would be successful?”

Outcome expectations and student’s social support:

Regardless of whether derived from their direct and indirect environment, they were measured by factor analysis of thirteen different statements in which the responders were asked to specify their level of agreement or disagreement on the basis of a four-level Likert-type scale 1—I disagree, 2—I probably disagree, 3—I probably agree, 4—I agree. The purpose of the analysis was to categorize the statements which were related and to find the uncountable quantities which are indicated by each factor.

Comments: The question, “Regarding your school performance, what is the area of study you would be most interested to study in higher education” (see question No. 4 of Appendix A) with possible answers: 1—Humanitarian sciences (such as languages, literature, arts, etc.); 2—Social sciences (such as sociology, political studies, economics, psychology, etc.); 3—Business and administration sciences; 4—Law, 5—Medicine, 6—Medical Support professional education (such as Nursing); 7—Natural Sciences and Mathematics (such as Physical and Biological sciences, Mathematics, Statistics, Computer Science, etc.); 8—Agriculture; 9—Engineering Sciences; 10—Architecture; 11—Other”, was modified as follows:

(i) Studies in higher education in non-STEM fields (answers 1-2-3-4-11)

(ii) Studies in higher education in STEM fields (answers 5-6-7-8-9-10).

Then, the second category (ii) was divided into two sub-categories: (a) Medicine (answers 5–6) and (b) higher education studies in other STEM fields (answers 7-8-9-10). As a result of this classification, the following three categories of students emerged: (1) Studies in higher education in non-STEM fields (answers 1-2-3-4-11), (2) higher education studies in Medicine (answers 5–6), and (3) Higher education studies in other STEM fields (answers 7-8-9-10).

It should be noted at this point that even though Medicine and Medical support professional education belong to STEM fields branches, in our research they are examined separately due to the fact that in these branches the social mission of the profession is accomplished and, thus, attract a large number of girls.

Statistical Analysis

The qualitative variables used in this study were expressed as absolute and relative frequency in each category of the variable. A Pearson chi-squared test (χ² test) was performed to assess possible differences that may exist in the questionnaire questions, in relation to the gender and students’ family economic situation. Furthermore, factor analysis was applied, in order to investigate the existence of common factors in a specific set of questions. In this group of new variables, the Cronbach credibility factor α was applied (Cronbach’s alpha). Finally, logarithmic regression was applied to explore the factors likely to be related to the student’s decision to pursue or not studies in STEM fields. The statistical analysis was carried out in the statistical program IBM SPSS Statistics (version 21). All the statistical checks were performed at the statistical significance level of 5%.

4. Results

This part demonstrates the results obtained from the analysis of the questionnaires. Initially, the descriptive analysis of the participants’ demographic characteristics was carried out and then followed the appropriate statistical analysis to investigate the research question.

Table 2. Descriptive characteristics of students (N = 301).

	N (%)
Gender
Boy	151 (50.17%)
Girl	150 (49.83%)
Education level
8th grade	126 (41.86%)
9th grade	85 (28.24%)
10th grade
Father’s educational level	90 (29.90%)
Non-academic education	135 (44.85%)
Academic education	166 (55.15%)
Mother’s educational level
Non-academic education	140 (46.51%)
Academic education	161 (53.49%)
Do you have older siblings
Yes	151 (50.17%)
No	150 (49.83%)
If yes, do they study at a higher education institution
Yes	110 (72.85%)
No	41 (27.15%)
Students’ perception of the family economic background
Low	62 (21.38%)
Average	216 (74.48%)
High	12 (4.14%)
How important is it for you to continue tertiary education
It is not important	11 (3.65%)
It is important	290 (96.35%)
Are you interested to study in an academic institution
Non-STEM field	105 (34.88%)
STEM field	180 (59.80%)
No	16 (5.32%)
If you were going to follow studies in higher education in a field that you would not be interested in, do you believe that these studies would be successful?
I wouldn’t be able to succeed in my studies	178 (59.14%)
I would be able to succeed in my studies	123 (40.86%)
In case you decide to pursue higher education, will your family support you financially during your studies?
Yes	231 (76.74%)
No	12 (3.99%)
I am not sure	58 (19.27%)
In case you decide to pursue higher education, will you be required to support financially your family during your studies?
Yes	40 (13.29%)
No	123 (40.86%)
I am not sure	138 (45.85%)
Amongst your close circle of friends and family, who in your view are most likely to influence your decision to pursue or not to pursue higher education?
My parents	193 (64.12%)
Relatives	19 (6.31%)
Teachers	20 (6.64%)
School counsellor	10 (3.33%)
Friends	32 (10.63%)
Other	27 (8.97%)
Do you know any people (in your close surroundings-family and friends) who hold an academic degree (do not count teachers and administration in your school)?
Yes	281 (93.36%)
No	20 (6.64%)
Have you ever visited an academic institution?
Yes, a STEM institution	141 (46.84%)
Yes, a non-STEM institution	38 (12.63%)
No	12 2(40.53%)
Extra-curricular activities
Yes, STEM	109 (36.21%)
Yes, Non- STEM	68 (22.59%)
No	124 (41.20%)
Learning difficulties in Mathematics
Not at all–Little	164 (54.49%)
Moderate–Very much	116 (38.54%)
I do not attend these lessons	21 (6.98%)
Learning difficulties in Computers
Not at all–Little	137 (45.51%)
Moderate–Very much	80 (26.58%)
I do not attend these lessons	84 (27.91%)
Learning difficulties in Physics
Not at all–Little	179 (59.47%)
Moderate–Very much	107 (35.55%)
I do not attend these lessons	15 (4.98%)
Learning difficulties in Biology
Not at all–Little	198 (65.78%)
Moderate–Very much	78 (25.91%)
I do not attend these lessons	25 (8.31%)
Learning difficulties in Chemistry
Not at all–Little	174 (57.81%)
Moderate–Very much	106 (35.22%)
I do not attend these lessons	21 (6.98%)
What is your average score in Mathematics
Low	33 (11.79%)
Average	93 (33.21%)
High	154 (55.0%)
What is your average score in Computers
Low	12 (5.53%)
Average	46 (21.20%)
High	159 (73.27%)
What is your average score in Physics
Low	23 (8.04%)
Average	95 (33.22%)
High	168 (58.74%)
What is your average score in Biology
Low	14 (5.07%)
Average	50 (18.11%)
High	212 (76.81%)
What is your average score in Chemistry
Low	23 (8.21%)
Average	77 (27.5%)
High	180 (64.29%)
Do you take private lessons in Mathematics
Yes, on a weekly basis	170 (56.48%)
Yes, but less often than once a week	13 (4.32%)
No	118 (39.20%)
Do you take private lessons in Computers
Yes, on a weekly basis	20 (6.64%)
Yes, but less often than once a week	4 (1.33%)
No	277 (92.03%)
Do you take private lessons in the following subjects (Physics, Chemistry, Biology)
Yes, on a weekly basis	167 (55.48%)
Yes, but less often than once a week	14 (4.65%)
No	120 (39.87%)

exhibits the distribution (N, %) of participants’ individual characteristics.

A factor analysis with Varimax rotation was then performed for question No. 6 of Appendix A: “Please express the level of agreement or disagreement with the following statements”, which includes 13 statements. The students were asked to answer a 4-level Likert scale with possible answers: 0 “I disagree”, 1 “I probably disagree”, 2 “I probably agree”, 3 “I agree”. The purpose of factor analysis is to categorize the statements that are related and to find the unmeasured quantities indicated by each factor.

Table 3. Factor analysis results for attitudes towards higher education.

	Factor Loadings	Cronbach’s Alpha
Factor 1: Social value		0.62
School should encourage students to pursue higher education through courses or programs	0.70
Everyone who is capable should pursue higher education	0.52
I would be proud to be studying at a higher education Institution.	0.47
It is important to acquire higher education	0.38
It is important for me to have satisfactory school performance so that I can pursue higher education.	0.31
Factor 2: Motivation movement		0.57
People with higher education credentials are more successful in life.	0.73
People with higher education have distinctly higher social status in Greek society.	0.55
People with a higher education degree are likely to practice higher-earned professions than those who do not have a higher education degree.	0.44
Factor 3: Social support		0.52
Most of my teachers and school administrators encourage me to pursue higher education.	0.75
My parents (or at least one of them) encourage me to pursue higher education.	0.44

illustrates the results of the factor analysis.

Based on the results of factor analysis, it appears that the data are suitable for applying factor analysis, as the statistical KMO gets a value of 0.755, which is considered satisfactory. In addition, according to Bartlett’s sphericity test (p < 0.001), there are significant correlations between the items, in order to extract representative factors for all the statements. Table 3 indicates that three factors were obtained from this analysis. It also demonstrates the factor loadings of each statement with the factor to which it belongs, as well as Cronbach’s alpha coefficient for each subset of statements, which were created according to the factor analysis. The first factor refers to the social value of higher education studies and the second on the motivation movement. These two factors reflect the outcome expectations. The third factor represents the perceived social support for academic education.

In addition, Cronbach’s alpha coefficient for the first factor, which refers to the social value of higher education, is 0.62, suggesting a satisfactory reliability of this factor. For the second and third factors, Cronbach’s alpha coefficient was 0.57 and 0.52 respectively, indicating less satisfactory reliability of these factors. It is worth noting that among the 13 statements of the 6th question in the above table, only ten are included. The remaining three did not show significant factor loadings and were therefore not included.

Considering the findings illustrated in

Table 4. Chi-square test for the evaluation of possible differences according to gender.

	Boy	Girl	p
Regarding your school performance and grades are you planning to pursue higher education			0.028
STEM fields	95 (67.9%)	80 (55.2%)
Non-STEM fields	45 (32.1%)	65 (44.8%)
Regarding your school performance what is the area of study are you most interested to study in higher education			<0.001
Medicine	17 (12.1%)	37 (25.5%)
Rest STEM fields	78 (55.7%)	43 (29.7%)
Non- STEM fields	45 (32.1%)	65 (44.8%)
If you were going to follow studies in higher education in a field that you would not be interested in, do you believe that these studies would be successful?			0.023
I would not succeed in my studies	77 (51.0%)	101 (67.4%)
I would succeed in my studies	74 (49.0%)	49 (32.7%)

, Chi-square test shows that there is a tendency for boys to be more interested in dealing with STEM fields. It is worth mentioning that 67.9% of boys choose a STEM field when the corresponding rate for girls is 55.2% (p = 0.028). Furthermore, it could be noted that there was an increased rate of interest of boys for STEM fields unrelated to Medicine (55.7% for boys, 29.7% for girls), along with a high percentage of girls versus boys dealing with Medicine and Medical Support education professions (25.5% for girls, 12.1% for boys) (p < 0.001). Finally, boys believe that they will succeed in their studies, even if they are not really interested in the field they choose to study (p = 0).

Additionally, Chi-square test shows that higher parental education has a positive impact on students’ decision to continue studying in higher education, but without specifying the field (p = 0.001 for father’s educational level, p = 0.011 for mother’s educational level) (

Table 5. Chi-square test for the evaluation of possible differences according to parental education.

	Regarding Your School Performance and Grades Are You Planning to Pursue Higher Education
	Definitely Not	Definitely Yes	p
Father’s educational level			0.011
Non-academic education	12 (75%)	123 (43.2%)
Academic education	4 (25%)	162 (56.8%)
Mother’s educational level			0.011
Non-academic education	11 (68.9%)	129 (45.3%)
Academic education	5 (31.3%)	156 (54.7%)

).

Finally, a multi-factorial logarithmic regression model was applied (logistic regression), which contained all the above factors, aiming at developing a prediction model for students’ response to higher education in STEM fields. The results are presented in

Table 6. Logistic regression analysis for students’ interest to study at STEM academic institutions.

Variables		OR *	p
Gender	Boy (cat. reference)
	Girl	0.886	0.831
Father’s educational level	Non-academic degree (cat. reference)
	Academic degree (cat. reference)	0.769	0.703
Mother’s educational level	Non-academic degree (cat. reference)
	Academic degree (cat. reference)	0.698	0.552
Student’s perception of the family economic background	Low (cat. reference)		0.069
	Average	0.198	0.034
	High	0.090	0.079
Self-efficacy	I am confident (for sure) I would not be able to succeed in my studies (cat. reference)		0.034
	I probably wouldn’t be able to succeed in my studies	7.002	0.014
	I would probably be able to succeed in my studies	14.366	0.005
	I am confident (for sure) I would be able to succeed in my studies	8.084	0.077
Extra-curricular activities	Yes, Non-STEM activities (cat. reference)		0.020
	Yes, STEM activities	8.166	0.010
	No	1.570	0.502
Have you ever visited an academic institution	Yes, a STEM institution (cat. reference)		0.086
	Yes, a non- STEM institution	0.091	0.029
	No	0.495	0.216
Do you take private lessons in Mathematics	Yes, on a weekly basis (cat. reference)		0.037
	Yes, but less often than once a week	0.521	0.710
	No	6.860	0.014
Do you take private lessons in Computers	Yes, on a weekly basis (cat. reference)		0.154
	Yes, but less often than once a week	44.559	0.179
	No	8.272	0.063
Do you take private lessons in the following subjects (Physics, Chemistry, Biology)	Yes, on a weekly basis (cat. reference)		0.002
	Yes, but less often than once a week	0.031	0.051
	No	0.071	0.002
Learning difficulties in Mathematics	Not at all–Little (cat. reference)
	Moderate–Very much	0.203	0.038
Learning difficulties in Computers	Not at all–Little (cat. reference)
	Moderate–Very much	2.598	0.178
Learning difficulties in Physics	Not at all–Little (cat. reference)
	Moderate–Very much	0.544	0.403
Learning difficulties in Biology	Not at all–Little (cat. reference)
	Moderate–Very much	0.509	0.311
Learning difficulties in Chemistry	Not at all–Little (cat. reference)
	Moderate–Very much	0.397	0.226
What is your average score in Mathematics	Low (cat. reference)		0.379
	Average	0.503	0.541
	High	0.183	0.194
What is your average score in Computers	Low (cat. reference)		0.453
	Average	4.349	0.259
	High	2.105	0.549
What is your average score in Physics	Low (cat. reference)		0.128
	Average	2.754	0.458
	High	11.592	0.089
What is your average score in Biology	Low (cat. reference)		0.309
	Average	0.073	0.129
	High	0.078	0.147
What is your average score in Chemistry	Low (cat. reference)		0.053
	Average	12.202	0.112
	High	1.609	0.754
Social value		0.607	0.279
Motivational movement		2.695	0.018
Social support		1.480	0.338

* OR = Odds Ratio, p < 0.05.

.

According to this model, it appears that the factors that predict the dependent variable (STEM field selection in tertiary education) are the following five: the financial situation of the students’ family, self-efficacy, the learning experiences, the difficulty in mathematics, and the motivational movement.

Specifically, children who consider their family to be in a middle economic situation, have a reduced probability of 80.2% of choosing STEM fields in tertiary education than those whose families are in a lower financial state (OR = 0.198, p = 0.034). It is clarified that the low and the low to medium economic situation has been considered as a low economic situation. Regarding the self-efficacy, students who responded that “I would probably not succeed in these studies” (OR = 7.002, p = 0.014) and “I would probably succeed in these studies“(OR = 14.366, p = 0.005) are more likely to choose STEM fields in tertiary education in relation to those who responded “I am sure I would not succeed in these studies”.

Students’ exposure to STEM activities includes visits to higher education STEM institutions, students’ involvement in STEM programs and actions both in and outside classroom, private lessons in Maths and Computer and Natural Sciences. In particular, students who had participated in STEM activities have a greater chance of choosing STEM fields in tertiary education (OR = 8.166, p = 0.010). Furthermore, it was noticed that students who had visited a higher educational institution, which does not belong to a STEM field, were more likely to choose STEM fields in tertiary education in relation to those who have visited a STEM higher educational institution (OR = 0.091, p = 0.029).

Private courses were also correlated with the option of choosing STEM studies, with students who take private lessons in Natural Sciences (Physics, Chemistry, Biology) less often than once a week (OR = 0.031, p = 0.051), or who had not experienced private courses in Natural Sciences (OR = 0.071, p = 0.002) being less likely to follow STEM academic studies than those who had. Furthermore, students who had difficulty attending Maths (moderate to very much) had decreased chances of choosing STEM fields in higher education in contrast to those who had little or no difficulty (OR = 0.203, p = 0.038). Finally, students who believe that higher education gives them higher financial profits, social status, and a “successful” social life are more likely to choose STEM studies (OR = 2.695, p = 0.018).

5. Discussion-Conclusions

Based on the statistical analysis of the findings of this study, it appears that gender factor does not play a regulatory role in determining the expression of interest for STEM studies. Although literature data suggest that there is an under-representation of girls in dealing with these fields, this is not confirmed by this research.

Girls are under-represented in Polytechnic Schools and Schools of Sciences, where at Medical Schools the girls’ attendance rate is higher than that of boys. The great interest of girls’ preference for medicine eliminates the difference of under-representation in the remaining STEM fields and therefore gender is not a regulatory factor, as SCCT claims. However, this is not impressive if we take into consideration that girls have a high degree of self-efficacy for schools in which their social mission is carried out, and Medicine is a STEM field of social character and a strong element of offering to fellow human beings [40,41].

It is worth mentioning that in the Competition PISA of 2015, which takes place every three years under the auspices of OECD (Organization for Economic Cooperation and Development), with a main focus on Natural Sciences and Mathematics-based STEM courses, Greek students aged 15 years have exhibited the same performance in the above-mentioned subjects regardless of gender. Furthermore, it is noted that in Greece there are no significant differences in the performance of boys and girls in Physics and Mathematics. Certainly, the almost identical performance of boys and girls in the above subjects does not indicate their preference for the field of study they will deal with, but it is an indicator of their abilities and skills. The latter is based on the Socio-Cognitive Career Theory combined with environmental factors and learning experiences that contribute to the development of the cognitive mechanisms of the individual and from there on to the manifestation of interest. We should never omit that the person is dealing with those activities in which he/she has a high degree of self-efficacy.

Additionally, the low percentage of students who do not wish to continue their studies in higher education does not create a database for comparative study with those that will proceed with STEM and non-STEM studies.

Although gender factor does not play a regulatory role that determines the expression of interest for STEM studies, the significance is that our findings support that boys have a high level of self-efficacy perceptions both for traditional males’ and females’ work, while female students have only for female occupations [12,33].

In terms of students’ perception of their economic family situation, it was revealed that students who are in low and low to average economic situations are more likely to be engaged in STEM studies. This result is in accordance with data that suggest that economically disadvantaged students are more likely to choose STEM fields for their studies [55], as well as from studies by George-Jackson and Lichtenberg, [56] and Lichtenberg and George-Jackson [57], which indicate that economically poor students have more confidence to deal with STEM fields. Τhe survey by Chachashvili-Bolotin et al. [58] proves that the higher the student’s economic family status is, the lower their interest in studying in STEM fields in tertiary education. In other words, economically disadvantaged chose STEM fields, keeping all other variables constant. This significant result is also confirmed by most recent research that suggests that economic disadvantages are related to STEM [55,56,57,58].

It has also been illustrated that students choosing STEM fields have a greater degree of self-efficacy than students who choose non-STEM fields. They are more confident about their abilities so they can successfully carry out a task, a finding that is in agreement with the relevant research findings, namely that people dealing with STEM fields are more confident about their capabilities for these fields [56,57].

The results of our research confirm literature studies [59,60,61,63], which relate the student’s exposure to STEM actions and programs both within and outside the school with an interest in STEM studies in higher education. Students who have visited STEM academic institutions or participated in STEM groups are statistically more likely to choose STEM studies in higher education. Increased attendance of STEM courses and STEM training increases the chances of students’ interest in STEM fields, reaffirming the principles of the Socio-Cognitive Career Theory [25,30,31].

Data of this study indicate that the student’s exposure to extra classroom STEM activities contributes to interest enhancement towards future studies of cognitive subjects of the same spectrum. It seems that during their visit to such institutions, students are often taught lessons in science of increased difficulty, and through activities they adopt a scientific way of thinking [61]. Indeed, the sooner the person is exposed to such activities, the more likely they are to study in STEM fields [59,62,63,64], confirming the significance of our study.

Furthermore, this is in agreement with relative studies that indicate that students’ performance and achievements in mathematics is one of the strongest predicting factors for further STEM studies [25] and it is confirmed for the first time in our research.

Finally, the mobility lever is a predictor variable for students interested in studying STEM in higher education. Research data indicate that children of low socioeconomic status choose occupations based on more external features, while students from higher social and economic status present a more internalized view of the choice of study and by extension, their profession. These external characteristics include remuneration, professional opportunities, basically the criteria of the mobilization lever [54]. It has already been mentioned that economically deprived choose studies based on the expected results to which the mobilization lever belongs [55]. The originality of this study is also based on this tendency.

Overall, in this research, learning experiences, student’s exposure to STEM activities at the school environment and outside of it, self-efficacy, and outcome expectations were found to be positively related to the procedure of students’ interest development to pursue STEM fields in higher education. On the contrary, the parental educational background only supports the student’s decision to continue studies after high school, while students from low-income families are more likely to follow STEM studies. Finally, gender seems to be a non-regulatory factor. It is worth mentioning that in this research most of the examined factors supported the SCCT.

6. Limitations and Suggestions for Future Study

A “missing point” of the questionnaire used in this research could be an inclusion of a question concerning the professional activity of the students’ parents. Perhaps we could have then drawn valuable conclusions about the fields of study in which students are interested, since there seems to be a correlation between father’s profession and student’s study field. This could be the subject of a study in a future survey.

Additionally, the research sample could be enlarged and not taken only from the city of Piraeus. In this case, we would also take into account the factor of the place of residence throughout the whole country. Another interesting feature in our survey could have been the participation of students studying at technical high schools, in which students receive a different educational level and practice. This could be an additional field for prospective research.