Digital learning resource use among Swedish medical students: insights from a nationwide survey | BMC Medical Education

Survey participants and response rate

Basic characteristics of survey respondents are provided in Table 1.

The response rate varied significantly between universities (12.3% – 28.0%; Lund 383/1370 = 28.0%, Göteborg 181/1474 = 12.3%, Linköping 242/1187 = 20.4%, Örebro 191/689 = 27.7%, Stockholm 213/1645 = 12.9%, Uppsala 311/1121 = 27.7%, Umeå 245/1236 = 19.8%). Total response rate across all seven sites was 1766/8722 = 20.2%. The sample was representative according to sex-distribution for the six out of seven sites that provided pertinent data (% females [% females at site]): Lund 62.4% [58.8%], Göteborg 60.2% [not reported], Linköping 62.8% [60.2%], Örebro 66.0% [65.0%], Stockholm 55.9% [54.5%], Uppsala 70.1% [64.6%], Umeå 66.5% [60.0%].

Student distribution across semesters can be found in Supplemental Figure 1. A majority of students attended clinical semesters, i.e., semesters 5–11 (n = 1035, 58.6%), whereas 731 attended preclinical semesters, i.e., semester 1–4 (41.4%).

Approximately four out of five study participants (n = 1426, 80.7%) responded to the final free-text answer question, providing vast qualitative data, ranging from three words to 360 words.

Which digital resources are used?

Frequency of use of different digital resources is shown in Fig. 1. The five most used digital resources were (percentage at least on a weekly basis): university digital study platform (75.3%), video (68.0%), digital flashcards (66.4%) digital student notes (53.4%), and external digital study platform (47.3%). Differences in the use of digital resources between the seven universities can be seen in Supplemental Figure 2. Usage frequency of several modalities varied substantially between universities, for example, at least weekly use of video (55.0% [Uppsala] – 81.9% [Umeå]), flashcards (51.3% [Örebro] – 84.0% [Lund]) and generative AI (29.8% [Lund] – 50.4% [Linköping]).

Frequency of use of different digital resources

Qualitative data (free text format responses to the final survey question) highlighted frequent use of multiple different specific digital resources including flashcard programs (e.g., Anki, Quizlet), generative AI (e.g., Chat GPT), YouTube channels (e.g., Ninja Nerd), external study platforms (e.g., Hypocampus, Osmosis), webpages (e.g., internetmedicin.se, Wikipedia), databases and search engines (e.g., PubMed, ClinicalKey, Google), and workspace and notebook applications (e.g., OneNote, Notion, Trello). A complete list of reported digital resources can be found in Supplemental Table 1. An overview of thematic analysis of free-text responses is provided in Supplemental Table 2. Although select, illustrative quotes can be found in the text below, a full presentation of the qualitative findings will be provided in a separate, Swedish-language article.

Which background factors influence frequency of use of digital resources?

Subgroup analyses revealed significant differences in the frequency of use of digital resources stratified according to age, sex, having children, regular extracurricular work, and stage of medical program (Table 2). Most differences were found when comparing resource use for those attending preclinical vs. clinical semesters, with preclinical students reporting significantly higher use of videos, flashcards, social media groups, generative AI, student notes, and digital books (Table 2, Fig. 2). In multivariable logistic regression analyses, younger age (≤ 25 years) was strongly associated with more frequent use of flashcards and generative AI, but more infrequent use of digital articles and podcasts (Table 3), although very few overall reported using podcasts in relation to medical school (Table 2). Female sex was associated with less frequent use of generative AI and digital articles, but higher use of digital student notes and external digital platforms (Table 3). Having children at home was associated with more frequent use of videos and university digital platforms (Table 3).

Digital resource use for those attending preclinical vs. clinical semesters

How do medical students find their digital resources?

Most students (74.8%) reported finding their digital resources based on recommendations from more senior medical students, through online search engines (57.0%) or by recommendations from the course management (40.7%). Few (7.6%) indicated ‘other’ ways of finding resources, and only 4.8% reported selecting resources after watching ads while using other digital resources.

Is use of digital resources promoted?

More than half of students reported that use of digital resources was promoted by the course management group (often, 19.6%; sometimes, 44.7%), whereas a significant minority responded ‘rarely’ (22.9%) or ‘never’ (5.4%). The extent to which such promotion influences use of digital resources appears to vary widely (high, 12.2%; moderate 37.2%; low 24.6%, not at all, 17.2%; don´t know 8.7%).

In which situations and contexts do medical students most often use digital resources?

Results related to which situations and contexts digital resources are used in are shown in Fig. 3. Almost all medical students (96.9%) reported using digital resources during individual studies.

Situations and contexts in which digital resources are most frequently used by medical students

Perceived advantages and disadvantages of digital learning resources in general

For digital resources in general, the reported primary advantages were availability (i.e., being able to decide when and where to use [90.9%]), possibility to pause/repeat [80.6%], more effective learning compared to other traditional learning methods [59.0%], updated/relevant content [49.0%], and ability to multi-task [17.3%]. The most common reported disadvantages were being easily distracted [49.6%], unreliable content [45.4%], difficulties separating study time and leisure time [40.5%], and difficulties finding useful content [21.6%]. Fewer than 10% reported that they found digital study resources less effective compared to other traditional methods [7.0%], or that the number of ads or commercial breaks was burdensome [8.7%]. Almost one out of five students were unable to identify any specific disadvantages associated with digital resources.

Perceived advantages and disadvantages of educational videos

Perceptions related to educational videos were specifically probed. Similar to digital resources in general, the two most common advantages were the possibility to pause/replay [89.8%] and availability (being able to decide when and where to use [84.8%]). Moreover, more than half [55.0%] found videos to be an effective way to quickly acquire concentrated knowledge, as well as prepare for upcoming classroom activities/lectures [50.5%]. Forty-one percent reported that videos may add a dimension that is hard to capture with other learning modalities and 33.0% found videos to provide more effective learning compared to other traditional methods. Only 2.8% reported not using videos for learning. Although only 1.1% were unable to see any advantages of video-based learning, several disadvantages were reported: being easily distracted [44.1%], difficulties judging whether the video content is applicable in a Swedish context [43.4%], difficulties finding videos with the desired learning content [31.9%], subpar quality [29.3%], unreliable information [22.1%] and too many ads/commercial breaks [10.5%].

The main reasons for discontinuing a video prematurely were as follows: does not fulfill learning needs [54.1%], message not sufficiently focused [36.0%], too low tempo [33.0%], presenter/lecturer not engaging enough [24.3%], subpar sound quality [19.1%], the program does not set aside enough time to view the recommended video material [17.6%], content does not match the title [9.6%] and subpar image quality [6.6%].

Qualitative data provided by 46 participants highlighted a few areas that were not captured by the survey questions:

• content is sometimes outdated, especially the universities´ material

• students sometimes watch the videos at a playback speed > 1 × to optimize their use of limited study time

• the quality and content of professional YouTube videos (certain creators are mentioned several times, e.g., Ninja Nerd) are often perceived markedly better than the universities´ videos: “…you get access to the best lecturers on the planet.” (participant #257)

• some students wish that teachers/lecturers could recommend, evaluate YouTube channels

• disadvantage with prerecorded video lectures: not being able to ask questions in real-time

• being able to watch video from home, even if you are sick

Perceived effect of simultaneous exercise on podcast-based learning

One question was devoted to exploring how medical students perceive the effects of exercise on learning while listening to podcasts. More than half of students [57.1%] checked the response option ‘I don´t listen to medical educational podcasts or don´t exercise’. For the remaining students, the perceived direction of the effect varied substantially: no effect [1.9%], very negative effect [4.9%], somewhat negative effect [7.8%], some positive effect [11.3%], very positive effect [3.7%], and 13.3% responding ‘don´t know’.

Flashcard-based learning: perceived effects and access to material

Flashcards were perceived to have a large to very large positive impact on development and maintenance of theoretical knowledge by 63.7% of medical students (Fig. 4).

Perceived effects of flashcard-based learning on development and maintenance of theoretical knowledge

There was an almost overwhelming praise of flashcards as a learning method, evidenced by extensive free-text answers from 32 participants, for example:

“I think Anki flashcards is the most effective way to learn. To create the cards you need to learn the subject, i.e., the creation itself becomes a learning process, and you can then repeatedly use the cards to learn.” (participant #54).

“Anki is the single most important factor in my studies.” (participant #946).

“Anki – cannot be overstated how much Anki helps medical students.” (participant #1064).

Several participants called for flashcards to be created by lecturers and course leaders, to accompany the educational material:

“It is a no-brainer that lecturers should hand out flashcards covering the content of the lecture, at least when we´re talking about anatomy and topics which demand intense memorization.” (participant #956).

“If I had one wish how digital resources could be used to improve medical school, it would be Anki-cards created and double-checked by the course leaders.” (participant #1731).

Flashcards were most commonly acquired from medical students attending higher semesters [65.7%], created by themselves [55.4%], or by fellow students attending the same course/semester [32.4%]. More rarely, flashcards were obtained from free online sources [21.5%] or generated de novo through specialized software, based on text/notes [5.9%].

Generative AI: reasons for using and frequency of use related to submission tasks

The main reasons for using generative AI were to search for information or find answers to questions [51.8%], summarize text [28.4%], find ideas to get started with tasks [24.0%], adjust/improve text [17.3%], receive feedback on texts [14.3%], and translate text [12.0%]. Students reported using generative AI for submission tasks relatively infrequently: never [41.3%], rarely [23.0%], sometimes [17.8%], often [11.9%], always [4.1%]. However, younger students (age ≤ 25 years) reported using AI sometimes/often/always for submission tasks more often than those > 25 years old (35.9% vs 30.7%, χ² = 3.9, p = 0.049).

Qualitative free-text data from 33 participants supported the wide utility of generative AI for medical students, and several students ranked generative AI (typically ChatGPT) as the single most useful digital resource.

Course literature: printed books and digital books

Relatively few medical students reported using (buying or borrowing) traditional printed course literature for every course [16.6%] or most courses [23.0%]. Most students reported using printed course literature for a few courses [36.6%]. Approximately one out of five students [22.7%] reported never using printed course literature. The main reasons for not using course literature were as follows: more efficient learning using other methods [36.6%], expensive [26.4%], not needed to fulfill learning objectives [16.9%], not being able to borrow the book [3.7%], too difficult/complicated [3.2%].

The frequency of use of digital books (e-books) is shown in Fig. 1. Ample qualitative data from 30 participants highlighted several additional points related to course literature:

• free access to online e-books is appreciated (for example, via university subscription to ClinicalKey)

• many students find the course literature list “unrealistic” – sometimes several extensive books are included even for brief courses

• lack of time is commonly indicated as a reason for not using the designated course literature

• printed books may be heavy and cumbersome to carry around

• e-books are preferred by many given the possibility to search for keywords, more easily navigate the chapters and copy images/text to student notes