Scholars of higher education policy have discussed a trend towards a practical, employability-guided curriculum within existing university programs, even traditional ones (Boden & Nedeva, 2010; Johansen et al., 2017; Madsen, 2022). These developments have, to our knowledge, not been studied systematically using quantitative text analysis on a nationwide dataset that allows describing the development of the description of higher education as more and more practical and applied. With such a study, we want to add systematic evidence on the development of self-descriptions of the curriculum. Hence, we ask the following research question:

Does the content of curriculum descriptions by Danish university increasingly highlight application and practice, and if so, how do these changes differ by fields of study?

Our main analysis covers the period from 2014 to 2021. The time frame is on the one hand limited by data availability. But it covers a period in which Danish universities were under significant pressure to increase graduate employability. Our findings show a nearly linear increase in the use of expressions that point towards a practical, applied orientation of the sector. This trend does not spare the humanities, which do not have a strong tradition of vocational orientation.

Overall, we test the following hypothesis:
Practical, applied and supposedly employability relevant wording increased its share of course descriptions over time. This increase is not only driven by a change in the composition of fields of studies, but also by changes within field (Hypothesis 1).

Based on theoretical assumptions, we derive two contradictory hypothesis on the dfferentiation of the development by fields of studies. From the background of subjects culture and identity as a place were students get practical and applied training, we would expect the following: 

Practical and applied wording increased its share of course descriptions constantly over time. Especially subject areas that are already known for their vocational orientation and good employment outlooks – such as Business and Technical Sciences – highlight this characteristic by using the respective wording to describe their curricula.(hypothesis 2a)

On the other hand, a policy was implemented in the time we observe that put enormous pressure on institutions and departments, whose graduates often have a high graduate unemployment. Especially shortly after graduation, these are often more traditional arts/humanities and basic sciences. Therefore, we could also expect that:

Increases in practical wording should be most noteworthy in programs that are usually comparatively far from vocational specificity and have a high graduate unemployment rate. This would affect especially affect the humanities, including classical humanities such as languages, history, philosophy or theology. (hypothesis 2b).

We test these two hypothesis by describing the change applied langauge in course descriptions over time.

Increases in practical wording should be most noteworthy in programs that are usually comparatively far from vocational specificity and have a high graduate unemployment rate. This would affect especially affect the humanities, including classical humanities such as languages, history, philosophy or theology.

We employ a large scale quantitative analysis of all course descriptions of the eight Danish universities. Data was retrieved from several web archives and databases and then combined. We analyze course descriptions from 2014 to 2021. In addition, we conducted analysis based on all retrievable course descriptions of sufficient quality for the period from 2004-2012. Due to the limited availability and the inconsistent format of the earlier course descriptions, we used archives from 2014 onwards for our main analysis. They confirm that the trend we observe in the newer data had started earlier. We measure the emphasis on practical and applied content in course descriptions using a curated lexicon, developed using a data-driven computer-assisted exploration method (Carlsen & Ralund, 2022). The method is designed to ensure an exhaustive search while curating terms for the lexicon using both clustering and similarity-based methods for search. As the course descriptions are both in Danish and English, we iteratively translated terms and used text sources in both languages in the computer assisted search process. We screened 532 unique terms from an iterative similarity search and went through 218 topics (+2000 words) located via the HSBM network clustering procedure (Gerlach et al., 2018). The process resulted in 131 unique terms (82 in Danish, 49 in English). To assess the validity of the curated lexicon, i.e. if it in fact counts terms used in practical and applied context, we compared the classifications to human judgement. For each university and for each term, we sampled 25 short contexts for validity assessment. Human coders labeled 8288 short contexts, judging whether they expressed an applied/practical or vocational focus or not. Furthermore, 512 labels were coded by both coders to assess inter-coder reliability. This resulted in an agreement percentage of 0.91, and a Kappa statistic of 0.64 (normalized by expected overlap), which demonstrates a decent quality of the labels. We then matched each course description to the lexicon, resulting in a count of application related terms for each course. To avoid conflating the increase in such terms over time with a change in the length of course descriptions, we normalized the count as a percentage of words. We describe trends across years and disciplines. For our sub-analysis by discipline, we use a classifier to identify fields of study of each course description.

We clearly confirm our hypothesis 1 and show that an increase in the use of practical, applied wording to describe course content is not limited to changes in programs offered but also occurs through changes in existing programs and even existing courses. Furthermore, when breaking down our analysis by broad fields of studies, increases appear also within fields that are initially low in the level of practical wording, but are often known for higher graduate unemployment, especially the arts/humanities. This points towards our hypothesis 2b, that the concrete political pressure to reduce admission was an important reason, and not just the general ‘zeitgeist’ towards applied higher education, which was the underlying assumption for hypothesis 2a. As a field that has initially resisted describing their content as applied, such changes are certainly more radical and further away from the traditional identity of this field. At the same time, humanities programs were hit hardest by the targeted reduction of subjects with difficult labor market outlooks for graduates. However, increases were nearly constant and linear over time, which also talks towards a longer lasting, more general trend as expected in hypothesis 2a. It is thus likely that both processes – the general debate and the concrete incentives – contributed to the pattern that we observe.

Boden, R., & Nedeva, M. (2010, 2010/01/01). Employing discourse: universities and graduate ‘employability’. Journal of Education Policy, 25(1), 37-54. Carlsen, H. B., & Ralund, S. (2022). Computational grounded theory revisited: From computer-led to computer-assisted text analysis. Big Data & Society, 9(1), 20539517221080146. Gerlach, M., Peixoto, T. P., & Altmann, E. G. (2018). A network approach to topic models. Science Advances, 4(7), eaaq1360. Johansen, U. V., Knudsen, F. B., Engelbrecht Kristoffersen, C., Stellfeld Rasmussen, J., Saaby Steffen, E., & Sund, K. J. (2017, 2017/02/01). Political discourse on higher education in Denmark: from enlightened citizen to homo economicus. Studies in Higher Education, 42(2), 264-277. Madsen, M. (2022). Competitive/comparative governance mechanisms beyond marketization: A refined concept of competition in education governance research. European Educational Research Journal, 21(1), 182-199.