Self-efficacy and (vocational) interest in technology and design: an empirical study in seventh and eighth-grade classrooms

A generally accepted theory in educational psychology holds that that interest is always directed towards an object, that is for example to specific topics, knowledge domains, activities, or goals (Krapp 1998; Krapp and Prenzel 2011). This implies that one cannot simply be interested but that one is always interested in something (see Gardner 1996). Moreover, interest determines the relationship between the person and the object of interest. Different types of appeal lead to an interest-oriented occupation with an object. Interest that is aroused by a cognitive, emotional, or value-related appeal in a specific situation is called “situational interest.” If an interest already exists because the object poses a cognitive challenge or arouses positive emotions, or because dealing with the object is regarded as a potential benefit, this is referred to as an individual interest (Krapp and Prenzel 2011). For operationalizing interest, cognitive, emotional, and value-related facets have to be taken into account (Krapp and Prenzel 2011). As they may vary with the same overall interests, also different activities, contents, and contexts pertaining to the object need be distinguished (Hoffmann et al. 1998; Todt 1995). For these reasons and because interests become more specific with increasing age (Daniels 2008; Krapp and Prenzel 2011), it is essential to go beyond global assessments like those of TIMSS and PISA (see, e.g., Martin et al. 2012; OECD 2007) where interests are considered without any particular focus.

A feeling that is closely linked to the feeling of competence, which is one of the basic psychological needs involved in the development of interests (Deci and Ryan 1993), is the feeling of self-efficacy (Bandura 1993, 1997, 2000). Self-efficacy is defined as “beliefs in one’s capabilities to organize and execute courses of action required to produce given attainments” (Bandura 1997, p. 3). The literature usually draws a distinction between a relatively stable general self-efficacy trait (Jerusalem and Hopf 2002; Schwarzer and Jerusalem 2002) and domain- or situation-specific self-efficacy that sensitively responds to personal developments and influences of the environment. The latter has often been used as an indicator of changes induced in teaching interventions (Bandura 1997; Zimmerman 2000).

According to the expectancy-value model proposed by Eccles and Wigfield (2002), expectations of success and subjective task values can be assumed to influence achievement-related choices and performance in a direct way (see Fig. 1). Expectations of success can be understood as an individuals’ beliefs about how well he or she will be able to cope with a given task, either in the immediate or long-term future. Such beliefs can be measured in a manner that analogous to measures of Bandura’s (1997) personal efficacy expectations. In the context of teaching and learning, Eccles and Wigfield (2002, p. 119) divide the subjective value attributed to a task or a topic area into four components:

The model assigns self-efficacy to the expectations-of-success component (Eccles and Wigfield 2002, pp. 110–111) while the various components of interest correspond to the subjective value of the task (Eccles and Wigfield 2002, pp. 114–115). The value-related component of interest can be assigned to the attainment value and the emotion-related component of interest to the intrinsic value. According to the model, sex and the gender-specific self-concept influence self-efficacy and interest via an individual’s perceptions, goals, and general self-schemata. These, in turn, influence career choice.

In the context of technology education, not many researchers have looked at the correlations and influences between interest, self-efficacy, and career choice or other constructs explained by Eccles and Wigfield (2002) although they predict choices of tasks and future jobs. Studies by Lawanto et al. (2012) as well as Lawanto and Stewardson (2013) were able to identify correlations between interest and expectations of success in engineering design tasks but their findings say nothing about causal relations.

The new Swiss curriculum “Lehrplan 21” (D-EDK 2016) makes references to technology education in four different subjects: “Technical and Textile Design,” “Nature and Technology,” “ICT and Media,” and “Economy, Labour and Housekeeping.” The decisive question is whether the intended interconnections between the different subjects, which are explicitly mentioned in Lehrplan 21, will indeed be adopted and put into effect in practice. If so, this could be a promising starting point for strengthening technology education in Switzerland in the next few years.

Primary school, where “Nature, Man, Society” is taught as an integral subject, would theoretically provide ideal conditions for discussing and approaching technology from a variety of perspectives, for example along the lines of the German concept of “multi-perspective technology education.” In reality, however, teachers do often not dare to deal with technical problems because on one hand they do not feel competent enough and on the other hand they do not have enough time for practical activities. Besides, the infrastructure of the classrooms does usually not allow handicraft and experimental and technology-oriented teaching. At lower secondary-school level, the combination subject “Nature, Man, Society” is split into four distinct subject areas, two of which with strong connections to technology: “Nature and Technology” and “Economy, Labour, and Housekeeping.” A comparison between the different domains of technology education in tertiary and vocational education and the corresponding school subjects relating to technology education according to Lehrplan 21 reveals that most domains are represented by a school subject except for “Technology and Engineering” (see Fig. 2). In the curriculum for upper secondary school, the subject “Technical and Textile Design” has been completely abolished, and science is taught separately in the three traditional, distinct main subjects “Physics,” “Chemistry,” and “Biology.” Thus, engineers and technicians are not involved in compulsory schooling in Switzerland, and the same applies to current teacher education programs.

According to the theoretical outline, the research findings and the situation in Switzerland presented above, it was of interest for us to analyze technology-related interests of adolescents by distinguishing between different areas of life (school, leisure time, and future job) and specific activities within a design process in general or in potential jobs. Moreover, interest and self-efficacy are personal-trait variables that are connected to each other and both influence vocational interests. Taking these premises into account, the three research questions to be answered in this paper are the following:

For measuring relatively stable individual interests and self-efficacy, self-report surveys have proved to be a useful data source and have therefore been used in many studies (e.g. Krapp and Prenzel 2011; Renninger and Hidi 2011). It is important, however, that the operationalization is carried out on the basis of clearly defined psychological constructs and that a variety of dimensions and facets of interest are covered.

In our study, we measured the participants’ individual interest in technical activities during a design process, their feelings of self-efficacy in solving technical tasks, and their vocational interests by means of a standardized questionnaire. It contained scales that had previously been tested and validated in large-scale assessments as well as scales whose reliability and validity had to be ascertained beforehand because of adaptations to the subject of technology. A pilot study using these scales was carried out with 102 pupils. A description of this pilot study and the scales is provided in Guedel (2014). All scales presented in this paper were 4-point Likert scales. The adapted, final version of the questionnaire was individually completed by 483 students, mostly online but in a few exceptional cases in a paper–pencil format.

The participants were seventh and eighth graders from all three academic tracks (I, II, III) of lower secondary school in Northwestern Switzerland. The sample was roughly equally distributed between girls (48%) and boys (52%), grades (7th grade: 45%; 8th grade: 55%) and the academic tracks (Track I: 44%; Track II: 34%; Track III: 22%). Since the track with the highest requirements (III) may seem to be underrepresented, it has to be noted that the actual population of this track is smaller than the populations of the low and the medium track (I and II).

The following section presents the results for research question 1: How do technology-specific interests of girls and boys differ between different contexts and different activities within a design process?

The following section presents the results for research question 2: How does self-efficacy in technical tasks as perceived by girls and boys differ between different tasks (irrespective of the content) within the design process?

The average of a total of 24 items relating to perceived technology-specific self-efficacy ranges from 2.33 (“Even if I am under time pressure, I can draw very accurate sketches and plans”) to 3.06 (“When I select a technical device, I always know what is important to me”). By means of principal component analyses, we could extract four factors from a total of 15 items (KMO = 0.89; MSA coefficients > 0.81; see “Appendix 3”). These factors explained 60% of the total variance. The mean values of the scales point to a medium perception of technology-specific self-efficacy (2.53 < M > 2.72). Overall, “Using and repairing” (M = 2.57) and “Planning and designing” received the lowest ratings (M = 2.52). In contrast to interest, self-efficacy is most pronounced with respect to the scale “Understanding and explaining” (M = 2.72). The differences between the scales are not large, however (d = 0.2–0.3).

In most technical activities, the girls perceived their self-efficacy much more negative than the boys, particularly regarding the category “Using and repairing” (d = 1.3; see Table 4). In this area, girls do not feel confident at all while they reported the same extent of self-efficacy as the boys regarding the category “Planning and designing”. These are also those activities that interest girls the most (see Table 1). Accordingly, the correlations between interest and self-efficacy in design processes proved to be high (r = 0.6).

The following section presents the results for research question 3: Can vocational interest in the field of technology and design be predicted by interest in technology and self-efficacy in technical tasks?

This question was answered on the basis of a structural equation model (see Fig. 3) that related to whether interest in dealing with technology in leisure time (i.e. subjective task value) or technology-specific self-efficacy (i.e. expectations of success) more strongly influences vocational interest (i.e. achievement-related choices/interest; see Fig. 1). The missings in the data of the total sample (approximately 2–5 per item or scale) were replaced by mean values because the omission of all cases with a missing in the 13 items or scales would have resulted in a considerable reduction (approximately N = 50) of the sample size.

If the model is calculated with the total sample (N = 483), the latent variables are well represented by the manifest variables with standardized weights or a factor loading of 0.50–0.96 and variances explained by the latent variables of 0.25–0.92. The fit indices of the model can be assigned to the good or even very good range, which applies to the overall sample as well as to the two groups separately (see Table 5). The fit indices alone do not say much about the quality of the model, however. An assessment of the quality requires a detailed analysis of the assumed relationships between and the influences of the latent variables.

The size of the standardized coefficients (see Fig. 4) suggests that the theoretically assumed relationships between the latent variables in the overall model are all relevant. Self-efficacy and interest can predict vocational interest in the overall model with 61% explained variance. The path coefficient (regression coefficient) between self-efficacy and vocational interest (0.64) is three times higher than the path coefficient between interest and vocational interest (0.20). The correlation between the two latent variables (“Interest” and “Self-efficacy”) is high (0.62). Owing to the marked content-related connections between some manifest scales or items, covariances between the associated errors are to be assumed. This applies to errors concerning the self-efficacy scales “Planning and designing” and “Craft-related tasks” as well as five items characterizing vocational interest. Direct correlations between manifest variables are not possible.

A group comparison (multigroup analysis following Airbuckle 2007) between girls and boys revealed that the model is valid for both groups (see Table 5). Total explained variance in vocational interest turned out to be smaller than in the total sample (girls: 43%, boys: 51%). The path coefficients connecting “Interest in technology in leisure time” and “Vocational interest in technology” differ significantly between the two groups: in the case of the girls, interest has a weaker influence on vocational interest (0.13) than in the case of the boys (0.29, see Fig. 5) and is not significant.

The data mirrors the traditional stereotypical picture: the boys in our sample are clearly more interested in technology than the girls, especially in the context of the future job. Thus, looking at technology-specific interest in more detail did not change the common overall picture. Although there are some activities that attracted the girls’ interest so that the gender gap did not become manifest (“Planning and designing” and “Designing eco-friendly products”), it seems in general that as soon as the word “technical” occurs, girls feel less attracted than boys. With respect to the design of the study this could indicate that the results might have looked differently if the formulation of the items had avoided “technical” altogether and if the word had been replaced by alternatives like “devices” or “products.” What is nevertheless surprising is that both girls and boys felt much less attracted by “Inventing, developing, constructing and building” in connection with the future job than without mention of a specific context. Apparently, being actively involved in technical activities—in crafts, for instance—is considered to be fun, but still adolescents cannot imagine occupying themselves with such activities on a professional basis. One explanation for this finding could be that these activities are perceived as something very complex and difficult in the real “high-tech job world” so that only brains who studied electro engineering or informatics can cope with them. This assumption is not completely unfounded in the digital age but nevertheless somewhat alarming because the next generation (of the industrialized world) is not willing to take responsibility for current technological developments although they affect our (social) lives more than ever before. Another likely interpretation of this finding is that the “professional world” in general is something rather distant at this age, that is to say 1 or 2 years before career choice. This explanation points to the importance establishing real contacts between the professional world and school already right from the beginning of lower secondary-level schooling, for example by offering short internships, regular opportunities to gain some practical experience, insights into different job profiles, or excursions.

The results concerning the gender gap in terms of the participants’ interest in different activities in the future job indicate that girls seem to have quite well-defined ideas of the contexts in which they wish “to deal with people.” As our analyses show, they do not want to do so in a technical context, because otherwise their ratings of the items “Selling technical tools” or “Developing new tools in a team” would have been much more positive. This low extent of interest in technical activities irrespective of the activity as such (selling, explaining, designing, inventing, etc.) suggests that the vocational interests of girls are more strongly shaped by the context “future job” or by the working sector than by the concrete activity itself. Besides, we know from pertinent literature that in the eyes of females, the image of technical professions does not correspond with the notion of an ideal profession (see acatech and VDI 2009).

As our analyses have shown, self-efficacy is a crucial factor of self-assessments of technology related abilities, which is consistent with Bandura’s (1997) work on self-efficacy beliefs in learning. The findings of our study are obvious, especially with respect to gender differences, and raise the following question: how can the considerable differences between girls and boys that became apparent in the structural equation model be accounted for?

Overall, the simplified expectancy-value model (Fig. 1) based on Eccles and Wigfield (2002) could be empirically confirmed. This noteworthy insofar subfactors and subvariables had not been incorporated into our version of the model. The subjective value of the task, for example, was represented by only one scale, namely “Interest in the context of leisure time,” which includes the emotional and value-related facets of interest (see Krapp and Prenzel 2011) as well as the frequency of the activity. Thus, compared to the “task value” considered by Eccles and Wigfield (2002), the “cost” variable (opportunity costs) was missing. It may be connected with the frequency of activity, however. The factor “Expectations of success” was comprehensively integrated through the four scales covering technology-specific self-efficacy (see Fig. 3). The analysis of the entire sample pointed to an unequal influence of interest and self-efficacy on vocational interest. If, however, the model was examined for the subgroup of the girls, the model fit got worse so that the prediction of vocational interest was no longer satisfactory. This indicates that apart from interest and self-efficacy other variables, which have not been included in the model so far, may influence the vocational interests of girls. According to Eccles and Wigfield (2002), variables like cultural milieu, perceptions, goals, self-schemata, previous achievements, affective reactions, and memories have an indirect effect on achievement-related choices. As for girls, it seems likely that certain social factors also have direct implications. Still, the path coefficient between self-efficacy and vocational interest is comparatively high (0.57) whereas the path coefficient between interest and vocational interest disappeared almost completely (0.13). From this we can conclude that girls tend to rule out a technical career for emotional or value-related reasons that are in line with their expectations of success in solving technical problems.

These findings lead to the following fundamental question: how can technology-specific self-efficacy be built up and promoted? Bandura (1997), the originator of the concept of self-efficacy, discussed four sources of self-efficacy: (1) mastering difficult situations; (2) observation of models; (3) social support; (4) perception of one’s own emotional excitement. If individuals do not experience these sources, they can hardly develop feelings of self-efficacy. By linking this theory and our results to the usual teaching practices in the science and/or technical design classroom, we can provide teachers with information on how important it is to foster the four sources of self-efficacy especially in girls. Another way of increasing feelings of self-efficacy and thus of encouraging interest in technology consists in providing technology education on a long-term basis. Acatech and VDI (2009), for instance, were able to show a positive effect of continuous technical teaching in Germany: “If the students are offered adequate technology education—as in some of Germany’s federal states—interest in technology is … higher” (p. 30, own translation). MoMoTech (Acatech 2011) even states that promoting technology at school alone is not enough and consequently recommends establishing networks between schools and out-of-school facilities. For achieving sustainable effects, such initiatives should ideally already include kindergarten and preschool. Starting with technology-related classes and programs in lower secondary school is too late. According to Erikson (1982), middle childhood is a particularly influential time for interventions because this is when behavioral habits that are critical to health and competence become entrenched and set and when skills that form the basis for personal identities and self-esteem are acquired. Furthermore, Simpkins et al. (2006) found that individual differences in self-beliefs and task beliefs develop in the course of the elementary school years and then are steadily being refined in response to feedback on performance and in processes of identity formation during adolescence. Against this background, it seems to be highly advisable or even vital to introduce multidimensional technology education already in elementary school, that is to say at the age when self-beliefs and task beliefs start to develop.

As many studies have shown and teachers know, it is quite hard to arouse young people’s interest in technology via theoretical approaches. Therefore, alternative ways of accessing the world of technology need to be looked for. Our results indicate that integrating practical, theoretical, creative, and social aspects of technology can make the subject attractive to many adolescents. By including product design as one part of technology education, several aspects can be connected, thus offering students a unique and special learning experience. In our study, we could observe that building cognitive bridges between practical work (manual skills) and theoretical contexts (knowledge of science) poses a great challenge for teachers as well as for students (Guedel 2014). Thus, one central question is how it is possible to bring these different activities (practical, theoretical, creative, social) together in such a way that it is not only fun, but the students also learn something? And what can they learn?

Lewis et al. (1998) suggested that the focus of technology teaching needs to shift from a design process or problem-solving approach to an approach that fosters problem posing. This approach is especially interesting to design teachers. Among the advantages Lewis et al. (1998) see the promotion of active participation and an increased likelihood that the teaching is characterized by authenticity and creativity. As children need to understand the situation from which the problem arises (Lewis et al. 1998), they should be allowed enough time to immerse themselves in the context of the task before the specific need, opportunity, or problem is presented or addressed. Further suggestions for successful learning in technology education have been proposed by Mawson (2003, p. 123): “The task itself should be sufficiently open to provide the possibility of a range of solutions being developed and the children would be encouraged to choose a starting strategy which fitted their own preferred learning styles. As designing and making occur concurrently, at regular intervals children should be required to explain/show, and discuss their developing solutions.”

Another approach, which is of special interest particularly to science teachers, has been outlined by Kim and Roth (2016, p. 153): “To integrate the societal emphasis on engineering and technology in the school curriculum, science teachers are encouraged to integrate in their curricula engineering and technological design tasks such as designing and building structures (i.e., boats, bridges, or devices).” This is supported by a meta-analysis on context-based science education, showing very large effects for societal contexts (STS = Science and technology in Society; Bennett et al. 2007).

Such recommendations often rest on those dimensions that science and technology share rather than on those dimensions in which they differ (Roth 2001). This is why there is a tendency to prioritize science content knowledge and skills acquired in hands-on technological design (Dohn 2013). In other words, within the framework of this approach, teachers regard classroom activities that relate to technology and engineering as a means for extending their students’ knowledge of science and accordingly tend to focus on how certain scientific concepts can be applied and verified in technological design. Kim and Roth (2016, p. 153) draw a similar conclusion: “There has been much emphasis on enhancing students’ science content knowledge and the application of science in technology and engineering education.” Although such efforts are to be welcomed in principle and can be regarded as a first step towards the aim of making students familiar with the world of technology, they are not sufficient to do justice to technology as a distinct domain that is worth being dealt with for its own sake and not just as a subordinate or “junior” field of the natural sciences. Hence, Kim and Roth (2016, p. 153) add: “…, it is important to remember the different goals of the two disciplines, and that they are both depending on each-other but have different goals.”