Development of a RIASEC-based model to analyse students’ perceptions of the work of nuclear- and astrophysicists

There are many stereotypical beliefs about the work of scientists among students. These can lead to a lack of interest and ill-considered career choices. This paper aims to shed more light on the everyday work of researchers in physics. To this end, an interview study (N=12) and a questionnaire study (N=66) were conducted among physicists in a Collaborative Research Centre for nuclear- and astrophysics in order to determine the physicists’ activity profiles using an adapted RIASEC+N model based on the theory of vocational choices by Holland. As a result, activities from all seven RIASEC+N dimensions were found, highlighting the broad spectrum of activities in modern physics research. However, with respect to the significance of each dimension, we found that the activity profiles of doctoral students and professors in general, as well as their profiles in theoretical and experimental groups differ particularly within the Realistic, Conventional and Investigative dimension. In the future, these results will be used to develop an outreach programme to give students an authentic insight into modern physics. Additionally, the activity profiles could serve as a reference for addressing stereotypes in school.


Introduction
In many countries there is an ongoing shortage of skilled professionals, which has even increased due to the Covid 19 pandemic depending on the work area.For instance, only in Germany, there were 300000 vacancies for specialists in mathematics, (computer-) science and technology in October 2022 [1].Against this background of the unmet need for professionals, particularly glaring in the STEM field, it is worrying that many students hold stereotypical ideas about the work of scientists [2], because these ideas (in addition to numerous influencing factors such as skills and personal goals) can be key factors for their career choices [3][4][5].
While many measures have already been taken to counteract these stereotypical ideas and to attract more students, especially more female students to science (for example with the pervasive establishment of traditional student laboratories), we argue that these laboratories must be complemented because they often focus on the experimental side and leave a blind spot for the work of theorists.Thereby they even run the risk of unintentionally reinforcing misconceptions.

Students' perceptions of the activities of scientists
When asked to picture a scientist, as it has long been done in the Draw a Scientist (DAST) tests [2, 6&7], students tend to illustrate rather stereotypical situations and people.They often think of scientists as old, sometimes crazy and mad men with glasses working in a lab alone [8].Besides the perceptions of external characteristics of scientists, such as clothing, gender or appearance, young people also show many stereotypical views of their professional activities.Stamer et al. [9] and Leiß [2] were able to show that some students have strongly abbreviated views of scientists' activities.Leiß uses an adapted DAST test in which typical work situations are to be drawn.As the most common idea students draw an experimental situation in 77% of the cases.Situations in which scientists work with theories are far behind as the second most common idea with only 11%.In about 87% of the cases, the scientists are pictured to work alone, dramatically underestimating or even overlooking the great importance of lively exchange, (international) collaboration and communication between colleagues.In addition to that the extent to which administrative activities play a role is hardly represented in the students' perception [2&9].At large it seems that many students hold stereotypical beliefs about the work of scientists which must be assumed to negatively affect their considerations of a science career.Moreover, a more elaborate knowledge of science related vocational fields is positively associated with [10] or could even have a direct influence [11] on students' career expectations.
However, so far most of the studies have asked for the perception of science in general, while there are only few studies about the perception of individual disciplines.With regard to physics, Heine and Pospiech have found that the theoretical, mathematical side of physics seems to be hardly represented in students' perceptions [12], but the database on this is sparse.Therefore, it is important to further assess students' perceptions of the job profile of physics as well as their interest and their perceived self-efficacy in the associated activities in the first place [13].

RIASEC+N model
This study is based on Holland's hexagonal RIASEC theory of vocational personalities and work environments.It states that each profession can be classified into one of the six RIASEC dimensions, namely Realistic, Investigative, Artistic, Social, Enterprising and Conventional.Each dimension is dominated by similar typical activities.For example, the Realistic dimension is dominated by manual activities such as building or repairing, while the Conventional dimension is characterized by administrative and repetitive tasks (see table 1).[14] Table 1.RIASEC dimensions with typical activities and professions [14].

Conventional administration Librarian
In addition, the model defines personality types by describing each person's vocational interests with a three-letter code, the order of which indicates a gradation of the person's interests.The link between work environment and personality types offers the opportunity to make promising predictions of vocational choices.So, if someone is strongly interested in Artistic activities, but still has an affinity for Realistic and Social activities, this person will be characterised with the personality type ARS.
Accordingly, the RIASEC model will indicate that working in a music school might be an advisable option [14].
Table 1 shows, that in general Holland classifies physicists or scientists into the Investigative dimension.Nevertheless, as more recent studies reveal, this does not represent the versatile tasks of scientist nowadays.So in a study about students' interests in science, Dierks et al. adapted the theory and found that scientists' work profiles comprise activities from all six RIASEC dimensions.In their study, they even found a new dimension called Networking, which includes activities such as peer-topeer communication or attending conferences [15].The now seven-dimensional RIASEC+N model has been used and validated in several studies to investigate students' interest in science and school science activities [16 -19].Wentorf et al. [13], Stamer et al. [9] and Leiß [2] focused more on students' perceptions of and interest in real science activities and aimed to break down their stereotypes using outof-school labs.To this end, Stamer et al. have conducted an interview and a questionnaire study with scientists in the first place, to get an authentic insight in the professional activities of scientists.This study confirmed the coverage of all seven RIASEC+N dimensions and showed a more multi-faceted picture, stressing that the activities of doctoral researchers are clearly distinct from those of group leaders and professors in the extent to which activities from each dimension occur.[9] However, in summarising all scientists, we assume that these studies still leave a blind spot for the specific work of physicists and especially theoretical physicists, with the focus being very much on the experimental side so far.

Research goals
Against this background, the aim of this study is to further specify the existing RIASEC+N model for the work of physicists and specifically researchers in the Collaborative Research Centre 1245 (CRC 1245) [20] in the field of nuclear and astrophysics at the Technical University of Darmstadt.Additionally, we want to prove, if the adapted RIASEC+N questionnaire is a suitable option to assess the activities of physicists.We hypothesise that there is a large variation in the activities of professors, post-doctoral researchers and PhD-students from theoretical and experimental groups covering all seven RIASEC+N dimensions.Therefore, we pose the following research questions: 1.
To what extent do the RIASEC+N dimensions reflect the activities of physicists in the CRC 1245? 2.
What activity profiles can be found for physicists in the CRC 1245?

General Approach
To achieve these research goals, we opted for a two-study approach.In a first step (addressing research question 1), we conducted an exploratory interview study to survey the specific activities of physicists, especially theoretical physicists, in the CRC 1245 (N=12) as broadly as possible.The guided interviews were oriented along the RIASEC+N dimensions, but were otherwise designed to be as open-ended as possible [21].A total of 46 different activities was found, of which only about 50% appeared in previous studies (compare [2&9]).In a second step (addressing research question 2), these activities were sorted into the seven RIASEC+N dimensions via expert rating (N=6) and then transferred into items for a questionnaire for all members of the CRC 1245 in order to subsequently determine activity profiles.In this paper we will focus on the second research goal and the questionnaire study accordingly.

Sample selection
All participants of the interview and the questionnaire study are members of the CRC 1245.The CRC 1245 has about 100 members including professors, post-doctoral researchers und PhD-students respectively from theoretical and experimental physics.In table 2 the number of participants of the questionnaire study (Q) as well as the sum for all subgroups (total) are shown.In our sample we excluded Master-and Bachelor-students, as they are less experienced and not employed at the university.
Because only one research scientist participated in the surveys and gave similar answers to all other post-doctoral researchers, both groups were combined for the analysis.

Instrument and Procedures
The web-based questionnaire was implemented on SoSciSurvey and contained 46 items representing the different activities that were revealed in the preceding interview study (see table A1).The participants were asked to rate how often they perform the 46 different activities during their work routine using Single Item Measures with a five-point rating scale from "never" (1) to "frequently" (daily) (5).Namely, they were asked to rate statements like "One of my tasks as a physicist is... to program new software for theoretical calculations".Table 3 summarises the number of items in each RIASEC+N dimension with an example.We also gave the opportunity to add activities in a free-text field if any were missing.The questionnaire was distributed to all members of the CRC 1245 in April/May 2022.Networking …to give a presentation to other physicists.5 In order to analyse the differences between the professions, we tested the data for significance.Since the data do not show homogeneity of variance and are not normally distributed, we used a pairwise Mann-Whitney U test in R with alpha error accumulation (Bonferroni) and calculated the effect sizes via Pearson's r [22].

Activity profiles of theoretical and experimental physicists
The mean values in table 4 show, that at least experimental physicists carry out activities of all seven RIASEC+N dimensions.They achieve the highest value in the Realistic dimension with  = 3.07,  = 0.81, while they rate activities in the Enterprising dimension lowest with  = 2.26,  = 0.67.In contrast, theoretical physicists seem to have hardly any activities in the Realistic dimension.They rate it lowest with only  = 1.19,  = 0.42.(Note that a rating of 1 means that they don't carry out these activities at all.) Activities within the Investigative dimension are rated highest with  = 3.28,  = 0.45 by theoretical physicists.Upon first reading their low rating in the Conventional dimension might be surprising, but seems perfectly reasonable with some experiment related items in this dimension (see section 5.3).Despite the relatively large differences in the Realistic and the Conventional dimension, there are no relevant distinctions in the remaining dimensions.Overall, it is noteworthy that the standard deviations are relatively large, especially in the Realistic, Social and Enterprising dimensions.To address these facts, we looked at the professions in more detail.

Activity profiles for different professions
In figure 1 and 2 the mean values for every profession is plotted over the seven RIASEC+N dimensions.The horizontal bars indicate significant differences between groups and are marked with the effect sizes.
Figure 1 shows the activity profiles for experimental physicists in the CRC 1245.You can see that professors and PhD-students, in particular, have significantly different activity profiles with large effect sizes.Interestingly, professors of experimental physics do not seem to be as involved in activities in the Realistic dimension as PhD-students, but have higher values in the Social and Enterprising dimensions.
Post-doctoral researchers differ from PhD-students only in the Social dimension.Figure 2 shows the activity profiles of the theoretical physicists.Overall, they have lower values in the Realistic and Conventional dimension than experimental physicists.Looking at the different professions, the higher rating of professors in the Social and Enterprising dimension is also evident here.In addition, professors perform activities in the Artistic and Conventional dimensions significantly more often than PhD-students.post-doctoral researchers show no significant difference to PhD-students with the exception of the Investigative dimension with a medium effect size.

Activity profiles within the RIASEC+N dimensions
In order to gain a deeper understanding of the activities of the different professions in experimental and theoretical physics, we now take a closer look at the dimensions on single item level in figure 3 and 4. Since there is hardly any significant difference between post-doctoral researchers and the other two professions, we only distinguish between professors (solid) and PhD-students (dotted) here.In Figure 3 it is noticeable that the four groups within the Investigative dimension have some activities in common, such as interpreting data, learning and reading literature, but also differ in other activities.On the right-hand side, you will find some items that are more frequently performed by theoretical physicists.Here, programming tasks dominate, such as investigating physical processes through computer simulations, calculating equations on the computer or software programming in general.At the top you will find some experimental activities such as planning experiments or using computer software for experiments.At the bottom are some specific tasks of professors such as writing scientific literature or reviewing literature or proposals from fellow physicists.Although the overall mean values of theorists and experimentalists are similar in the Investigative dimension, they differ quite a bit at the level of the individual items.
Figure 4 provides more insight into the Conventional dimension.You can see that there are three items, that are related to experiments, such as repetitive settings on experiments, writing proposals for experiments or learning about occupational health and safety.
All other dimensions show a similar distribution on single item level and the mean values of professors and PhD-students mostly differ only in quantity.

Discussion and limitations
Our results suggest that the adapted RIASEC+N model is suitable to describe the specific occupational activities of physicists in the CRC 1245.Our questionnaire with the 46 items found in the preceding interviews seems to depict their work adequately, as hardly any activities were added by the participants in the foreseen free-text field.The items span over all seven dimensions; in particular there are also items falling within the new Networking dimension.Thus, we can confirm the results of previous studies that activities within the Networking dimension are important for the work of physicists.Nevertheless, it should be noted that the Investigative dimension is still the dimension with the most items and high ratings, therefore Holland's original sorting cannot be completely discarded.
As we could complement the experimental items with many programming-related items within the Investigative dimension, we believe that we could refine the instrument significantly in order to capture activities in theoretical physics more adequately besides the activities in experimental physics.
Contrary to our hypothesis, post-doctoral researchers and PhD-students have similar activity profiles, with the exception of the Social dimension.The biggest difference is between professors and PhDstudents and here mainly in the Enterprising and Social dimension.This reveals that professors are in duty of nearly all management tasks and are responsible for most teaching activities such as lecturing or mentoring students.This fact, that there is such a large variation between professors and PhD-students respectively from theoretical and experimental physics indicates the need for differentiation in future studies.From our point of view, it will be necessary to open up the possibility of differentiation between those groups in the questionnaires used to assess the perception of the work of physicists, otherwise one would provoke that answers are always inappropriate for one group or the other.
With regard to the limitations we must keep several aspects in mind interpreting the results of this study.Since the participants only answered the questions on a rating scale from "never" to "frequently", as we do not expect anybody to track his or her working hours, we cannot make any statement about the actual frequency of execution.It is also not possible to rank the individual activities because the statements do not indicate how much time an activity takes.So if somebody rates two items with the same score you cannot tell which activity is more time-consuming.In addition, individually disliked activities can lead to an overestimation of the frequency.Some people might overrate the Conventional dimension because they think filling out formulars takes ages even though it only takes up a small part of the day.We tried to prevent this through the design of the questionnaire by adding specific time indications to the rating scale (e.g. ( 4) "often (weekly)").Furthermore, the assignment of the activity items to the dimensions is not fixed.Thus, a few of our items are found in other dimensions in other studies (e.g.Leiß sorts the activity "giving talks to peers" in the Enterprising dimension [2]).This can possibly lead to a shift in the frequency of the dimensions.

Conclusion and Outlook
With regard to the research goals, we were able to identify activity profiles of professors and PhDstudents in theoretical and experimental physics.We were able to show that their activities are similar in certain respects, but also differ in essential aspects such as experimental (Realistic dimension) or programming (Investigative dimension) related activities.For professors we see a focus on managing (Enterprising dimension) and teaching (Social dimension) tasks.The RIASEC+N model with the corresponding instrument in the adapted form presented here is useful to depict the activities of physicists in nuclear and astrophysics.
Based on the insights gained, we designed a digital 360° Serious Game and are planning a project week at TU Darmstadt in spring 2023 to give students an authentic insight into modern physics research and reduce their stereotypic beliefs about the work of physicists.The instrument together with the corresponding results can therefore also serve as a basis to record students' perceptions of the activities of researchers in physics in the future and as a reference while addressing stereotypes.

Figure 3 .
Figure 3. Mean ratings in the Investigative dimension on single item level for professors (solid) and PhD-students (dotted) respectively from theoretical (blue) and experimental (red) physics

Figure 4 .
Figure 4. Mean ratings in the Conventional dimension on single item level for professors (solid) and PhD-students (dotted) respectively from theoretical (blue) and experimental (red) physics

Table 2 .
Participants of the questionnaire study (Q) from all Members of CRC 1245 (total) respectively from theoretical and experimental physics (in spring 2022)[20]

Table 3 .
Questionnaire items in all RIASEC+N dimensions.

Table 4 .
Comparison of means: M and SD for RIASEC+N dimension