Effects of Image-based and Text-based Activities on Student Learning Outcomes

Anne K. Greenberg
CRLT University of Michigan 1071 Palmer Commons 100 Washtenaw Ave. Ann Arbor, MI 48109 +1 734-763-2880

Melissa Gross
Movement Science University of Michigan 3738 Central Campus Recreation Building 401 Washtenaw Ave. Ann Arbor, MI 48109 +1 734-764-9663

Mary C. Wright
CRLT University of Michigan 1071 Palmer Commons 100 Washtenaw Ave. Ann Arbor, MI 48109 +1 734-936-1135

akgreenb@umich.edu ABSTRACT

mgross@umich.edu

mcwright@umich.edu

Research on benets of visual learning has relied primarily on lecture-based pedagogy, not accounting for the processing time students need to make sense of both visual and verbal material[8]. In this study, we investigate the potential dierential eects of text-based and image-based student learning activities on student learning outcomes in a functional anatomy course. When controlling for demographics and prior GPA, participation in in-class image-based activities is signicantly correlated with performance on associated exam questions, while text-based engagement is not. Additionally, students rated activities as helpful for seeing images of key ideas and as being signicantly less mentally taxing than text-based activities.

Categories and Subject Descriptors

H.1.2 [User/Machine Systems]: Human FactorsHuman information processing ; J.3 [Life and Medical Sciences]: Health; K.3.1 [Computer Uses in Education]: Collaborative Learning

General Terms
Design, Measurement, Theory

Keywords
LectureTools, active learning, visualizer-verbalizer, dual coding, assessment

time, it was thought that a mix of verbal and visual formats would allow students to encode information in long-term memory twice[7, 2], thus increasing learning. Surprisingly, there is little evidence for this approach[3, 6]. For example, Tangen et al.[8] compare student quiz scores following three dierent types of lectures with visuals relevant to the content, with visuals not relevant, and with text-based bullet points and nd no signicant dierence in accuracy among them. Lin and Atkinson[5] nd similar results, albeit for animated vs. static graphics. However, research on visuals may be helpfully contextualized by two key factors. First, visual learning formats may be more benecial in course contexts requiring the interpretation of complex visual material [3, 4], such as anatomy or art history. Second, research on learning through visualverbal presentations has relied primarily on lecture-based pedagogy, not accounting for the processing time students need to make sense of both pieces of information[5, 8]. Therefore, we oer a important case study of student learning in an anatomy course at the University of Michigan. Anatomy is a discipline with a signicant visual component, for students to recognize and identify corporeal elements such as bones, ligaments, and muscles. Additionally, this course devotes signicant in-class time to active learning, which we hypothesize allows for the cognitive processing needed to enhance student learning. Previous studies have treated active learning as a separate condition from visual and text-based learning (e.g., Tangen et al.[8], and Bockhoven[1]), but we seek to investigate their integration.

1. INTRODUCTION
Many instructors seek the most eective way to present course material to students, often pairing visual formats (e.g., slide images) with verbal explanations. Indeed, at one

2.

METHODS

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for prot or commercial advantage and that copies bear this notice and the full citation on the rst page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specic permission and/or a fee. LAK 14, March 24 - 28 2014, Indianapolis, IN, USA. ACM 978-1-4503-2664-3/14/03$15.00. http://dx.doi.org/10.1145/2567574.2567597

The study analyzes data from a University of Michigan class, Human Musculoskeletal Anatomy, a sophomore-level course required for all students majoring in Movement Science. Taught fall and winter terms, the course averages 101 students, and it is comprised of about 60% Movement Science majors and 40% non-majors. The class provides students with an in-depth knowledge of the musculoskeletal system in order to better understand the functional relationships between human musculoskeletal anatomy and body movement. Each class session consists of lecture segments punctuated with active learning activities in which students work independently or in pair-share dyads to answer interactive ques-

Figure 1: Examples of in-class activities tions asked through LectureTools, interactive presentation software (Fig. 1). Both text-based and image-based activities are included in each class session. Text-based activities consist of a written question and multiple-choice answers. Image-based activities consist of either a question associated with a particular image (and multiple-choice answers) or an image map where students are asked to click on the image in order to answer the question. Student participation on all learning activities is tracked using LectureTools analytics data. The number of activities in which a student participated is collected for each student. Student learning outcomes are assessed via exam question scores. Individual exam questions are written to correspond to single lecture topics (and therefore either image-based or text-based learning activities). Care is taken to write exam questions of equivalent diculty. Regression analyses are used to look for correlations between student scores on exams and student participation in learning activities. Demographic data including gender, underrepresented minority status, and cumulative GPA are also included in analyses. Additionally, students were asked to participate in a endof-semester survey, which asked them to reect on how much mental activity (i.e., cognitive load) was required when participating in image- and text-based activities. Survey questions were modeled on the instrument for measuring cognitive load described by Lin & Atkinson[5]. Student response rate was 88%.

Table 1: Regression model predicting exam performance with in-class participation and demographic variables. R2 =0.418. URM status and gender are coded as dichotomous variables, with male=0 and female=1, and non-URM=0 and URM=1. Estim. SE p -value Intercept 34.6547 9.0839 0.0003 Image Participation 0.7844 0.2927 0.009 Text participation 0.2391 0.4090 0.56 URM status 4.6957 7.0671 0.51 Gender -13.7643 4.0102 0.0009 GPA 4.6383 2.0907 0.03

(based on GPA) are more likely to participate in text-based questions. This is not the case for image-based questions, indicating that participation in image-based in-class questions raises exam scores. The active learning classroom is an ideal setting for advancing our understanding of how text-based and imagebased materials support student learning. In the active learning setting, students have more time to process information, thus reducing cognitive load and potentially increasing deeper learning, particularly with image-based activities.

4.

REFERENCES

3. INITIAL FINDINGS AND DISCUSSION

We predicted that participation in all learning activities would be positively correlated with student-learning outcomes. However, given the visual nature of the discipline and active learning format, we hypothesized that participation in image-based learning activities would be more strongly correlated with exam scores than participation in text-based activities. Students reported that with active learning, image-based activities present less of a cognitive load. Nearly all students (95%) agreed or strongly agreed that LectureTools learning activities were helpful for seeing images of key ideas in the course. Additionally, on a scale of 1 to 8, with one corresponding to easy and 8 corresponding to demanding mental activity, students rated image-based activities as 4.14 (SD=1.83) and text-based activities as 4.78 (SD=1.81), a statistically signicant dierence (p <0.0001). Linear regressions suggest that participation in both textbased (r=0.46) and image-based (r=0.47) in-class activities are correlated with performance on associated exam questions (p <.0001). However, when GPA is taken into account using multivariate regression, participation on text-based questions is no longer signicantly correlated with exam grade (Table 1), indicating that higher performing students

[1] J. Bockoven. The pedagogical toolbox: computer-generated visual displays, classroom demonstration, and lecture. Psychological Reports, 94:967975, 2004. [2] J. M. Clark and A. Paivio. Dual coding theory and education. Educational Psychology Review, 3:149210, 1991. [3] B. Kolloel. Exploring the relation between visualizer-verbalizer cognitive styles and performance with visual or verbal learning material. Computers and Education, 58:697706, 2012. [4] S. H. Kotze, C. G. Mole, and L. M. Greyling. The translucent cadaver: an evaluation of the use of full body digital x-ray images and drawings in surface anatomy education. Anatomical Sciences Education, 5:287294, 2012. [5] L. Lin and R. K. Atkinson. Using animations and visual cueing to support learning of scientic concepts and processes. Computers and Education, 56:650658, 2011. [6] R. E. Mayer and L. J. Massa. Three facets of visual and verbal learners: Cognitive ability, cognitive style, and learning preference. Journal of Educational Psychology, 95:833846, 2003. [7] A. Paivio and K. Csapo. Picture superiority in free recall: imagery or dual coding? Cognitive Psychology, 5:176206, 1973. [8] J. M. Tangen, M. D. Constable, E. Durrant, C. Teeter, B. R. Beston, and J. A. Kim. The role of interest and images in slideware presentations. Computers and Education, 56:856872, 2011.