Introduction:

 

The Multimedia effect consists in combinations of text and visualizations that improve learning outcomes.

 

In the Process-oriented view: visualizations may help learning when learning depends on extracting and reasoning with visuo-spatial information, by enabling and facilitating cognitive processes that are less demanding than the one required when learning from text. 

 

According to the outcome-oriented view, redundant information presentations should aid learning compared to a complementary information thanks to the dual coding process. But this assumption contradicts the redundancy effect in multimedia learning. Learners have to coordinate the processing of text and pictures  in working memory. But this coordination is not necessary to end with a deeper understanding. Redundancy may lead to extraneous cognitive load that hinders learning. 

 

Here is a summary of the two approaches:

 

 

 

	Process-oriented view
	Sometimes	Yes
	When learning depends on extracting and reasoning with        visuo-spatial information.
	Cognitive process less demanding than the one used to learn from text.
	----
Democratized in literature?
	How information should be distributed across text and visualization? = focus of current study

 

 

 

 

 

Hypotheses:

 

 

1: The Outcome-oriented view leads to a stronger multimedia effect for text containing a high degree of spatial information.  Because in high spatial text, visualizations containing the same information can be better linked to the verbal information and dual coding more likely to occur.

 

2: The Process-oriented view leads to the exact opposite: strong multimedia effect for low spatial text. Because when representing little spatial information in the text, the information is distributed across representations in an optimal way for the understanding.

 

3: Moreover, according to the literature background, the multimedia learning is helpful when visualizations are complementary to the verbal information.

 

4: On a working memory level, a high spatial text may yield interference with the processing of the visualization: the text and the processing of visualizations demand visuo-spatial working memory resources.

 

5: A high spatial text can be redundant to the visualizations. Thus, there is an unnecessary coordination activities that triggers extraneous cognitive load. 

 

 

 

Method:

 

73 participants (age: M = 22,18; SD = 3,19 / 50 female, 23 male)
2 x 2 design with the degree of spatial information given in the text (high vs. low) and the presentation format (written text-only vs. written text + animation) as between-subjects factors. 

 

Procedure, 5 steps:

0) Prior knowledge questionnaire

Experiment, Time per session : 30 min.

1) Subjects were instructed to acquire as much information as possible from the multimedia instruction.

 

 

	HIGH SPACIAL TEXT	LOW SPACIAL TEXT
Written text only + Static picture	TEXT WITH SPATIAL INDICATORS                              +                   STATIC IMAGE	TEXT WITH NO/FEW SPATIAL                          INDICATORS +                  STATIC IMAGE
Written text + Animation	TEXT WITH SPATIAL INDICATORS                             +               ANIMATED IMAGE	TEXT WITH NO/FEW SPATIAL                          INDICATORS +               ANIMATED IMAGE

After having studied the materials, participants filled different test:

2) Cognitive load questionnaire.

3) A retention test and a visual test of 6 min. each + 3 minutes to comment on their drawings to facilitate scoring of ambiguous drawings.

4) Transfer questions for 6 minutes.

 

 

Measures:

 

The results have been evaluated through 5 tools: 

 

 

Prior knowledge	Cognitive load	Retention test	Transfer test	Visual test
Asses the general knowledge of the participant about the subject.            (thunderstorms).     7 Multiple choice items with 4 possible answer per item.     One credit per good answer. Scores from 0 to 7.	To asses how cognitively demanding the experiment was. (More specifically the overall load measure, intrinsic cognitive load, extraneous cognitive load, and germane cognitive load).     4 questions with a 7 point Likert scale (from 1 very easy/ pleasant / to 7 very difficult/ unpleasant).	To evaluate how much the subjects remembered they were asked to write down an explanation on how lightning works.     One point was given each time the participant could recall one of the 34 main idea of the lightning formation.	Transfer tests asses the adaptability aspects of performing changes related to learning. Transfer tests involves performing a previously learned task in a novel situation.     4 items with a six alternative multiple-choice. (3 being correct and 3 incorect).     One credit for each good answer Scores from 0 to 12.	The visual test is here a complement to transfer and retention test. It evaluates th ability of the participants to recall visual information on lightning formation.     5 items, One per sheet of paper representing a simplified background of the static picture.   For each appropriate and identifiably drawn visual element one point was gained.   Score 0 to 12.

 

 

 

Results: 

 

• Prior knowledge was about the same for every groups (about the same mean and SD: 1.95<M<2.40 (max. 7) and 0.78<SD<1.40)

 

• ANOVAs (between-subject factors. Cf. 2x2 matrix) did not reveal any main effect of spatial text information for retention, transfer and the visual test.

 

• However, as expected, significant multimedia effect for retention and for the visual test (significant probability p<.001) but not for the transfer test (p=0.06). Participants’ learning from written text with animation performed better than participants’ learning from text-only.

 

• No multimedia effect for high spatial test / strong multimedia effect for low spatial text

 

Discussion:

 

This article explores the interactions between the multimedia effect and the degree of spatial information conveyed through a text on learners.

 

Firstly most of the results seem to deliver proofs of the multimedia principle (you get deeper learning from words and picture than from pictures alone). But this multimedia effect is proved untrue, or at least not as true as expected in the high special text condition. These results show that former findings and theories about multimedia effect in learning might be inconsistent and inaccurate.

 

The general conclusion of the present study is that we need to explore more in depth the relationship between text and visualizations in multimedia learning and more specifically the conditions under which this effect is likely too occur. Indeed the present papers makes blurry the recent findings on the subject.

 

Eventhough the article gives more accurate information on this phenomenon some of this information tend to prove the multimedia principle untrue. Indeed the degree of spatial text information ( HIGH or LOW conditions) may influence  the participants' results  on the visual test indicators. More particularly concerning the internal representation of visuo-spatial information and, less strongly but still significantly, in the retention test.

Two explanations might be given to explain theses: (a) a capacity explanation for dealing with redundant information and (Cognitive Load Theory) and (b) an interference hypothesis in processing spatial information from different sources (Redundance theory).

These results tend to prove the outcome oriented theory wrong. Indeed this theory would predict an higher score in the high spatial text condition (Dual Coding effect).

 

On the other hand the results don’t prove the cognitive load theory to be true either. Indeed, the cognitive load measurements from the experiment show that a higher overall cognitive load was reported in those multimedia conditions with low rather than high spatial text.

 

Concerning further researches the article wonders which part exactly of the working memory are involved and potentially hindered by a text evoking processes of mental imagery and interfering with a process of visualizations. This question is still open and would maybe prove true the theory of Carretti et al. (2005); Pazzaglia & Cornoldi (1999); or the ones of Baddeley’s (2007).

Secondly the article evokes the interest of researches which would tackle the issue of eye movements behaviors and strategies. Indeed eye tracking experiments would very accurately deal with crucial assumption of the redundancy and the interference hypotheses:  text and animations are both attended. Eye tracking experiments would evaluate a potential split-attention effect (Scheiter & van Gog, 2009; Schmidt-Weigand et al., 2010; van Gog & Scheiter, 2010). For example, one may expect that learners attend to visualizations earlier, longer and/or more often when there is only little spatial information in the text.

 

 

 

 

 

References:

 

[1] Baddeley, A. (2007). Working memory, thought, and action. Oxford psychology series. Vol 45.

 

[2] Carretti, R., Borella, E., Cornoldi, C.,  De Beni, R. (2009). Role of working memory in explaining the performance of individuals with specific reading comprehension difficulties: A meta-analysis. Learning and Individual Differences. Volume 19, Issue 2: 246–251.

 

[3] Pazzagliaa, F., Cornoldia, C. (1999). The Role of Distinct Components of Visuo-spatial Working Memory in the Processing of Texts. Memory, Volume 7, Issue 1.

 

[4] Scheiter, K., van Gog, T. (2009). Using eye tracking in applied research to study and stimulate the processing of information from multi-representational sources.

 

[5] Schmidt-Weigand, F., Kohnertb, A., Glowalla, U. (2010). A closer look at split visual attention in system- and self-paced instruction in multimedia learning. Learning and Instruction, Volume 20, Issue 2: 100–110.

 

[6] Schmidt-Weigand, F., Scheiter, K. (2011). The role of spatial descriptions in learning from multimedia. Applied Cognitive Psychology, Volume 23, Issue 9: 1209–1214.

 

[7] Schmidt-Weigand, F., Scheiter, K. (2010). Eye tracking as a tool to study and enhance multimedia learning. Learning and Instruction, Volume 20, Issue 2: 95–99.