Decorative pictures and learning with media

Statistical concepts are quite abstract and difficult for students to understand, leading to high examination failure rates. This study investigated whether the use of decorative pictures during online learning could enhance learning performance or reduce cognitive load. Although many multimedia studies showed the negative influences of decorative pictures, the pictures enhanced learning if they induced learning‐relevant emotions and were integrated into the learning context. The purpose of this study was to examine the influence of decorative pictures on online statistical learning, using an online experiment during which students learned about nonparametric tests. The experiment included two sections, each with an emotional picture, learning materials, and questions. Retention and transfer tests were also carried out. Overall, 100 students participated in the experiment. A 2 × 2 between‐subjects design was used with the factors “picture in the first section” (positive vs. negative) and “picture in the second section” (positive vs. negative). The results showed that the first positive picture enhanced learning performance for the questions in the first section, while the second positive picture led to lower intrinsic and extraneous cognitive load; however, the experimental factors did not influence emotions, retention, or transfer. In summary, positive pictures in a learning context reduced cognitive load or enhanced learning, depending on their positions in the learning materials.

Learning materials often consist of pictures that do not pursue a learning goal but an aesthetics function. Such decorative pictures were found to hinder learning by an increase of learning-irrelevant cognitive load (i.e., the seductive detail effect). Studies also showed that various details within a learning setting could have a retrieval-enhancing function when shown again in a retrieval phase (i.e., the memory cues effect). This talk is aimed to present results from an experimental series with four experiments (with more than 100 participants in each experiment) testing decorative pictures in a learning video on marketing concepts as memory cues. To reach this effect, one group achieved these decorative pictures again in their learning tests. All experiments showed that the seductive detail effect occurs when pictures were only shown in the learning videos. Besides, when decorative pictures are also used in the learning tests, learning results increased in comparison to a group without decorative pictures. Across the experiments, significant moderators (e.g., the salience of the cues, the associative connection between the cues and the learning information, and the modality of the cues) were found. Results can be partially explained by differences in additional measures on mental load, mental effort, and intrinsic motivation.

The present study examines the effects of emotionally valenced pictures on learning outcomes in two different experiments (E1; E2). Each experiment’s objective is to reveal different aspects of positively and negatively valenced decorative pictures, like influences on arousal, irrelevant thinking, and motivation scores. Eighty-two (E1) and 121 (E2) university students had to learn facts about cell division (E1) or volleyball rules (E2) on different web pages and were randomly assigned to either a positive affective or negative affective decorative pictures group (E1), or positive active, negative active, and negative de-active pictures (E2). After finishing these environments, retention and transfer tasks had to be completed. Additionally, scales of pleasure, arousal, and attention (E1), or irrelevant thinking, and intrinsic motivation (E2) were implemented. Both experiments show that positive pictures are better suited to increase retention and transfer, although transfer performances might be equally improved through positive active and negative active pictures in contrast to negative de-active pictures. Positive pictures generate significantly higher scores of pleasure, arousal, and intrinsic motivation and lower scores of irrelevant thinking. Designers of instructional multimedia environments should be aware of these effects in order to achieve best learning performances.

Decorative pictures in multimedia materials were long been regarded as detrimental (seductive) for learning. In contrast, resent research has shown that the impact of these pictures is moderated by a number of variables. In a series of experiments, this presentation examines if the emotional charge of decorative illustrations can be verified as an additional moderator. The first experiment investigates the effectiveness of different levels of valence (positive vs. negative) shown in decorative pictures. Results show that learning performance is better for positive than negative pictures. In addition, positively charged pictures enhance the intrinsic motivation and pleasure of students. Arousal, which is defined as a second dimension of affect, was also influenced. Hence, a second experiment was conducted in order to examine if the dimension of arousal (also called activation) plays a role in the explanation of emotionally charged pictures. In this experiment, students were assigned to three decorative picture groups (positive activating, negative activating, and negative deactivating) of a multimedia learning environment. Results show that a positive valence and a higher arousal enhance retention performances. In contrast, transfer was only enhanced by more activating pictures. Results of a third experiment with an additionally included control group without decorative pictures reveal that positive activating decorative pictures can be seen as conducive and negative activating pictures as seductive for learning.

Instructional designer use external representations such as pictures in order to represent, organize, explain or decorate a learning material. In the case of decorative pictures (e.g., photographs or illustrations), the aim is to make a learning material aesthetically appealing rather than to provide learning topic-related information. As such, these pictures are frequently considered as extraneous materials that impede learning by overloading the working memory capacities of learners. However, the category of decorative pictures varies in a huge amount of distinct factors related to the arrangement of such pictures or their content. By reviewing current empirical findings, several boundary conditions can be identified. In this talk, four of these moderators will be presented by a series of experiments. Two experiments (N1 = 81, N2 =108) examine the effects of anthropomorphism, visible through human forms (with vs. without human faces) and the degree of personalization of picture labels (with vs. without personal addressing) for decorative pictures. Since the learning topic was about artificial intelligence, decorative pictures displayed robots in daily life but did not convey information according to the learning topic. Both experiments included measures on learning performance, as well as cognitive, affective and motivational impacts. Experiment 1 showed a significant increase of learning scores for both decorative pictures with human faces and highly personalized labels, while cognitive, motivational and affective measures significantly differed between groups. The second experiment additionally revealed learning advantages for anthropomorphic pictures (with human faces and personalized labels) in contrast to a control group without pictures. Based on cognitive-affective theories, the emotional charge and the degree of text-picture connectedness of decorative pictures will be shown as two other possible boundary conditions of decorative picture impacts. In order to test these moderators, three experiments (N3 = 108, N4 = 86, N5 = 162) with secondary (Experiment 3 and 5) or university (Experiment 4) students were conducted. For this, decorative pictures were tested in a 2 (positively vs. negatively charged) x 2 (weakly vs. strongly connected to the text) between-subjects design with an additional text-only condition. These pictures had been chosen consistent with a learning text about South Korea Experiment 3 and 4) or the human body (Experiment 5). Measures of affective responses, cognitive processes, and learning performance were included. Both moderators were found to influence the learning performance, while decorative pictures with a strong connection to the learning topic of a text and a positive charge were shown to increase learning. In comparison with a text-only condition, positive pictures enhanced learning, while weakly connected, negative pictures impaired learning. Results can be explained by differences among the assessment of cognitive processes of learners. In conclusion, decorative pictures may be used in order to foster learning if boundary conditions like the degree of anthropomorphism, connectedness, and emotional charge are taken into account.

Anthropomorphisierung bezeichnet im Lernkontext die Tendenz von Personen, menschliche Charakteristiken auf nicht-menschliche Objekte zu übertragen. Dieser Effekt wird in multimedialen Lernmaterialien durch Designelemente wie menschliche Gesichter oder soziale Hinweiszeichen hervorgerufen und kann die Lernleistung erhöhen. Bisher konnten Studien diese Tendenz nur für lernrelevante Bereiche in Lernmaterialien nachweisen. Dekorative Bilder, die als lernirrelevant definiert und häufig mit der Forschung zu seductive details verbunden werden, sind bisher nicht auf den Effekt der Anthropomorphisierung untersucht worden. In einem ersten Experiment (N=81) mit 2x2-faktoriellen Design wurden die Faktoren Einbindung von Gesichtern (ja vs. nein) und Personalisierung von Bildbeschriftungen (ja vs. nein) in vorher getesteten dekorativen Bildern in einem Lernmaterial über künstliche Intelligenz untersucht. Dabei wurden die Einflüsse dieser Bilder auf Lernergebnisse sowie kognitive, motivationale und emotionale Prozesse der Studierenden gemessen. Die Ergebnisse zeigen, dass dekorative Bilder mit Gesichtern (p<.001) oder einer personalisierten Beschriftung (p=.002) nicht nur die Lernergebnisse erhöhen, sondern auch die mentale Anstrengung (p=.001) von Lernenden steigern. Gesichter in Bildern verbessern außerdem die intrinsische Motivation (p=.003) sowie Valenz (p<.001). In einem zweiten Experiment (N=108) wurden anthropomorphisierende Bilder (mit Gesichtern und personalisierten Bildbeschriftungen) sowie nicht-anthropomorphisierende Bilder (ohne Gesichter und personalisierte Bildbeschriftungen) mit einer Kontrollgruppe ohne dekorative Bilder verglichen. Außerdem wurden die kognitive Belastung sowie motivationale und emotionale Zustände von Schülern gemessen. Dabei zeigt sich, dass anthropomorphisierende Bilder erneut die Lernergebnisse gegenüber nicht-anthropomorphisierenden Bildern (p<.001) aber auch gegenüber einer Kontrollgruppe (p=.002) verbessern. Obwohl sich die lernirrelevante Belastung von Lernenden in beiden Bildgruppen (p=.002) erhöht, steigern sich auch lernbezogene, kognitive Prozesse (p=.004) sowie die intrinsische Motivation (p=.001) und Valenz (p<.001) bei anthropomorphisierenden Bildern gegenüber beiden anderen Experimentalgruppen. Dieser Ergebnisse bestätigen, dass dekorative Bilder moderierenden Effekten unterliegen und eine Debatte über den Einsatz sowie die Beziehung zur seductive-detail-Forschung solcher Bilder neu bewertet werden muss.

When people attribute human characteristics to nonhuman objects they are amenable to anthropomorphism. For example, human faces or the insertion of personalized labels are found to trigger anthropomorphism. Two studies examine the effects of these features when included in decorative pictures in multimedia learning materials. In a first experiment, 81 university students were randomly assigned to 1 cell of a 2 (human faces vs. no faces in pictures) × 2 (personalized vs. nonpersonalized labels of pictures) between-subjects, factorial design. In addition to learning performance, cognitive, motivational, and emotional impacts of anthropomorphism are examined. Results show that both human faces and anthropomorphic labels were able to increase the learning performance on cognitive assessments. However, only human faces were able to influence motivational and emotional ratings significantly. In a second experiment, 108 secondary school students were randomly assigned to 3 groups (anthropomorphized pictures, nonanthropomorphized pictures, and no pictures) in order to evaluate possible advantages of anthropomorphism in decorative pictures in learning materials. Results show again that anthropomorphized pictures are better for learning than nonanthropomorphized pictures and also better than a control group. Results are discussed in the light of a debate on the inclusion or exclusion of decorative pictures.

Decorative pictures, which make a learning text aesthetically appealing rather than provide information, have been predominantly found to impair learning by an increase of learning-irrelevant cognitive processes. Recent research, however, indicates that this effect is moderated by various factors. On the basis of cognitive–affective theories and studies, the affective charge and the degree of text–picture connectedness (i.e., the semantic relation of text and pictures) of decorative pictures reveal possible boundary conditions. To examine these design features and compare them with a group without pictures, 3 experiments (N1 = 108; N2 = 86; N3 = 162) with secondary school (Experiments 1 and 3) or university (Experiment 2) students were conducted. For this, decorative pictures consistent with those in instructional texts about South Korea (Experiments 1 and 2) or the human body (Experiment 3), were tested in a 2 (positively vs. negatively charged) × 2 (weakly vs. strongly connected to the text) between-subjects design with an additional control group. Learning performance, affective responses, and cognitive processes were measured. Results show that students with either positive or strongly connected pictures outperformed students with negative or weakly connected pictures. In comparison with the control group, strongly connected positive pictures enhanced learning and weakly connected negative pictures impaired learning. Although negative pictures were shown to increase task-irrelevant thoughts and extraneous cognitive load, weakly connected pictures increased the perception of intrinsic cognitive load.