Professional Documents
Culture Documents
Amy B. Adcock
Cognitive load is generally defined as the amount of mental resources necessary for
information processing. High cognitive load requires the user to expend extra memory resources
in order to deal with incoming information. It is speculated that the necessity for extra resources
can cause a deficit in processing efficiency and social performance. Several empirical studies
link cognitive load effects to the components of working memory by examining the effects of
presentation modality on working memory load (Goolkasian 2000; Mayer & Moreno 1998;
Mayer, Moreno, Boire & Vagge 1999; Mousavi, Low & Sweller 1995).
Other tests of cognitive load theory focus on the modality of information presentation and
effects of split attention without relating these effects to the components of working memory
(David & Hirshman 1998; Kalyuga, Chandler & Sweller 1997; Velayo & Quirk; Yeung 1999).
Empirical tests have also investigated the role of cognitive load on types of performance
(Katsikopoulos, Duse-Anthony, Fisher & Duffy 2000; Lewis & Linder 1997; Pontari &
Schlenker 2000). The purpose of this paper is to examine research studying the effects of
cognitive load with an emphasis on the relationship between cognitive load and processing of
information.
“The term working memory refers to a brain system that provides temporary
storage and manipulation of the information necessary for such complex cognitive
tasks such as language comprehension, learning and reasoning” (Baddley, 1992).
In Baddley’s (1992) theory of working memory, he divides the memory system into three
separate components. The central executive component controls attention and oversees two slave
components. The phonological loop is responsible for processing speech-based information such
as spoken language. The visuospatial sketchpad processes visually depicted information such as
Effects of Cognitive Load 3
diagrams. Because, according to the theory, these two slave components act independently of
It has been hypothesized that cognitive load issues may be related to the components of
working memory. Several researchers have examined effects of cognitive load on information
retention based on this architecture of working memory (Goolkasian, 2000; Mayer & Moreno,
1998; Mayer, Moreno, Boire & Vagge 1999; Mousavi, et al., 1995). These studies have sought to
support the idea that information processing can occur in both slave components of working
Research on capacities of working memory components test for effects when subjects are
given concurrent input in different modalities. Several of these studies have also seen evidence
of a split attention effect, which occurs when subjects must divide attention between separate
tasks and mentally integrate sources of information. It is thought that this process of integration
increases cognitive load thereby decreasing performance and that this effect could also be
Researchers interested in examining the effect of presentation modality and split attention
have performed studies investigating how information processing is impacted by the amount of
cognitive load on working memory components. Mousavi, et al. (1995) investigated the
constructed using two different modalities (e.g. visual-auditory) to those presented with
information using the same modality (e.g. visual-visual). They speculated that a dual modality
presentation would decrease cognitive load and therefore increase working memory capacity.
This increase in capacity should manifest in better retention of the materials. Although they did
Effects of Cognitive Load 4
not directly measure cognitive load in this study, some interesting findings support existing
geometry. Each condition differed by modality of information regarding a geometry problem and
associated diagram. Some were presented with a visual diagram and text explanation, others
were allowed to hear a narrative explanation and look at a visual diagram. Retention of material
Their findings indicated that subjects performed better when presented with a dual
modality, i.e. visual diagram and auditory explanations. They deduced that this dual presentation
did decrease the cognitive load on working memory so that the subjects were able to process
both formats at the same time. More specifically they stated, “because both systems can be used
[that must be processed] is presented in a manner that permits it to be divided between the two
systems rather than processed in one system alone” (Mousavi, et al. 1995, p 331).
In this same study, researchers also looked at effects of split attention in terms of the
architecture of working memory (Mousavi, et al. 1995). Their interest here is to further support
the evidence that working memory components have the ability to simultaneously process
different modalities. The split attention effect is seen when subjects must divide their attention
process visual and auditory information at the same time then the necessity to mentally integrate
the two different modalities should disappear. This increase in capacity to focus on and process
Conditions varied in terms of how much time was allowed for subjects to study the
material and the presentation modalities. These modalities were identical to those described
above. The hypothesis for this part of the study is that the increase in working memory capacity
allowed through dual presentations modes will make integration of information easier by
reducing cognitive load. Quantifying the time taken in solving the problems was used as
evidence. Findings from the research did indicate that when subjects were required to mentally
Mayer and Moreno (1998) tested theories of cognitive load and working memory in
terms of a split attention effect. They believe the split attention effect is caused by subjects being
required to integrate and process multiple sources of information. They were interested in testing
were processed separately. They felt if visual and auditory components were presented
simultaneously, no split attention is necessary. Because they are being processed in two different
places there is no need to mentally integrate the information. Therefore, a lower cognitive load
would result and retention would be higher. Their study presented subjects with a multimedia
explanation of the causes of lightning. In one condition, subjects viewed a visual diagram and
text based explanation. The other condition viewed the diagram and heard the explanation as a
narrative.
those in the diagram-text condition. The authors concluded that when different sources of
material are presented in the same modalities, working memory is overloaded and deep
processing cannot occur. If information is presented in different modalities, subjects had more
Effects of Cognitive Load 6
room in their cognitive systems to hold the information. Also, because modalities are different,
mental integration is not necessary and cognitive load is reduced even more.
In a later paper, Mayer et al (1999) expanded on the concept of working memory load
and examined how multimedia environments can be used to help people integrate verbal and
visual information. Subjects were shown a multimedia instructional unit on auto mechanics.
Conditions varied by order of presentation and narration. Some were presented with concurrent
animation and narration, and some subjects received small or large “bites” of narration and then
animation. Subjects were tested for retention by several methods. Findings indicated that groups
presented with verbal and visual information concurrently outperformed the other conditions.
They concluded that learners were “more able to build referential connections between
corresponding visual and verbal representations when both are held in working memory
simultaneously” (p 643).
Other researchers have examined the split attention effect on components of working
memory and cognitive load. Goolkasian (2000) tested the how presentation format affects
reasoning capabilities. Her rationale for the study states that attendance to multiple sources of
information causes a split attention effect, which interferes with reasoning capacity. When
material is presented in formats that are easily integrated, the load on working memory is
reduced resulting in more efficient processing. She hypothesizes that participants will show best
formats, pictorial, textual or auditory. In some cases, formats were combined. After being shown
stimuli, participants were shown background information about the objects. They were then
asked a verification question about the presented material. Reaction times and error rates were
Effects of Cognitive Load 7
measured for all participants. In general, pictorial representations fared best in terms of reaction
times and error rates. However, the experimenter did see an unexpected format effect in the last
two experiments. Reaction times during problem solving were faster when participants were
required to integrate information over formats. She proposes these results could be due to a
reduction in cognitive load and an increase in working memory capacity and when integration
cognitive load and components of working memory. Each of the tests performed attempted to
draw a link between the defined components of working memory and the effects of information
presentation on their capabilities. It may make sense to relate the findings of these studies to the
limited capacities of the working memory system. These limitations, it appears, can be remedied
by carefully presenting information in a manner that does not overload one’s working memory
components through a singular format. Some researchers have studied the effects of split
attention and presentation modality without necessarily drawing links to Baddley’s working
memory theory. This paper will now examine the results of these studies.
The purpose here is to briefly review literature addressing recall performance and
cognitive load that does not directly refer to Baddley’s theory of working memory. Although it is
not directly referenced, an examination of these studies could be interpreted as being compatible
with the studies reviewed previously. These researchers were primarily interested in testing
instructional formats to see if presentation modality and split attention had any effect on
retention processes.
Effects of Cognitive Load 8
Presentation Modality Effect. Velayo and Quirk (2000) performed an experiment that
attempted to determine how modality influenced recall on a paired-associate learning task. They
were interested in supporting previous research showing improved recall and retention from
audiovisual presentations. Subjects either observed concept pairs through pictorial, textual or
auditory representations or were presented with mixed modalities. They were then tested for
mixed modalities. Although these results match those found in previously discussed studies,
these authors do not discuss the concept of working memory components as a justification for
their results. Instead, they relate problems associated with cognitive load and presentation
modality to difficulties in encoding. They concluded that the difficulties in encoding information
in similar modalities imposed a high cognitive load on subjects and had a negative effect on
recall.
Split Attention Effect and Levels of Expertise. Other studies have examined the role of
cognitive load management and effects of split attention and expertise in the subject area. Again,
split attention is defined as the process of mental integration of elements in order to make sense
of presented information. This mental integration is responsible for a large amount of cognitive
load and empirical testing seeks to reduce its effect on the cognitive system (Kalyuga, Chandler
& Sweller 1997; Yeung 1999). Yeung examined the effects of cognitive load by presenting
subjects with two vocabulary formats, a separated but integrated glossary. The purpose of the
study was to see if the process of attending to two distinct sources of information might impose a
high cognitive load through the split attention effect. Subjects at varying levels of expertise were
load by self-perception of the subjects. This is the only mention of direct measures of cognitive
Findings indicated interesting aspects of the split attention effect. For subjects with a low
amount of expertise, an integrated format that reduced the split attention effect proved more
effective for comprehension. High-level knowledge subjects performed better with a format of
separated terms. He speculates that the integrated format reduces cognitive load by eliminating
the need to search for meaning resulting in increased comprehension but the presence of the
meanings within the text increases redundancy. This redundancy effect has a negative impact on
Kalyuga, Chandler and Sweller (1997) also found that levels of expertise affect the
impact of cognitive load. The research tested whether the integration of text and diagrams
imposed a higher cognitive load negatively impacting retention. They wanted to examine
whether multiple elements of information would impose a higher cognitive load on subjects.
They felt that if a person is familiar with material, expressions are processed as single elements.
Therefore, multiple elements should not impose a high cognitive load on expert users. They
tested their hypotheses by testing physically integrated information against information that must
Findings indicated that the integrated diagram and text group outperformed the other
groups. They also found that the diagram and text group outperformed the text only group. They
concluded that this integrated format is effective for low level subjects but they did see a
negative redundancy effect for high-level learners. They suggest that this redundancy effect may
impose a higher cognitive load on subjects with a high level of expertise and negatively impact
their learning.
Effects of Cognitive Load 10
So far, the previously reviewed empirical studies have examined effects of cognitive load
on retention of materials. This is partially due to the fact that most of the research is centered in
the field of information processing. They all seem to indicate that a high cognitive load whether
caused by differing presentation modalities or by split attention effect has a negative impact on
retention of materials. Researchers have attempted to find ways of presenting information that
reduces the need for users to expend mental resources by presenting complementary modalities
and by eliminating the need to mentally integrate sources of information. One other interesting
line of research in this area has looked at cognitive load and social performance. The focus of
this paper will now turn to empirical studies examining the effects of cognitive load on self-
Several studies have looked at the role of cognitive load in performance effectiveness
(Katsikopoulos, et al. 2000; Lewis & Linder 1997; Pontari & Schlenker 2000). The general
consensus in this field of research is that increased cognitive load hampers performance. The
research that will be discussed in this paper manipulated cognitive load when doing different
Lewis and Linder (1997) sought to account for the effect of pressure on task
performance. The provision of a distraction is counted as the mechanism for adding cognitive
into the performance system by diverting attentional processes from the task. Subjects were
asked to putt a golf ball toward a target point. To add a cognitive load, subjects were given a
distraction task of counting backward from 100 by twos in an audible voice. Findings indicated
that the distraction task was sufficient to decrease performance significantly. However, when
pressure to perform was increased, performance was not significantly different from the
Effects of Cognitive Load 11
nondistracted condition. The authors explain this by stating if this pressure is consistent
throughout the mastery of the task, the attentional processes required during performance are
decreased and an increase of cognitive load will have no effect. They also feel that it would be
interesting to replicate this experiment with highly trained performers to see the resulting effect.
Pontari and Schlenker (2000) constructed empirical tests for cognitive load and self-
performance. They believe that previous research often starts with the idea that limited cognitive
resources make it more difficult for people to perform more than one “cognitively effortful task
at once” (p 1093). Experimenters asked subjects to role-play during an interview scenario. They
were not informed of their role type until after the experiment. Cognitive load was manipulated
by asking the participants in the experimental condition to recite a rehearsed eight-digit number
during the interview process. Later they were assessed on recall of the number and a comparison
Their findings indicated that self-presentation activities could span from mindless and
automatic to those requiring high cognitive resources. Subjects that defined themselves as
introverted seemed to need the highest amount of cognitive resources for self-performance. Their
manipulations of cognitive load impaired performance but only for introverted subjects playing
extroverted cognitive roles. Decidedly, the task of changing roles from an introvert to an
extrovert is a demanding one. Specifically they state, “in a social vacuum, cognitive busyness
logically leads to inferior performance on demanding tasks that require cognitive control”
Another study in the field of cognitive load looked at how participants alter decision-
making during driving. It was hoped this study would add to the field of traveler information.
Subjects were tested using a driving simulator, baseline data on route choice was provided by a
Effects of Cognitive Load 12
previous experiment. They speculated that drivers would have to eliminate at least one of the
route attributes (range of alternative route, time of travel) needed for decision making with an
increase in cognitive load. It is thought this would simplify the process of deciding on a travel
route to account for resources used to deal with the increased cognitive load. The findings from
their study supported their speculations showing that drivers consciously decreased their task
expectations when given a higher cognitive load. The authors here do state limitations to this
study such as methodology. Overall though, it seems that the results support previous findings
Conclusion
The research studies examined here were fairly consistent in their definitions of what
cognitive load is and how it affects human processing and performance. They all state the
resources available in human memory and how it can alter cognitive processes can define
cognitive load. In these studies, cognitive load was studied as a result of the effects it manifests.
There is also strong evidence presented linking effects of inducing cognitive load with Baddley’s
theory of working memory. Studies of presentation modality show how information presented in
a way that incorporates coding for individual modalities are effective at reducing cognitive load
and increasing retention. Research examining the split attention effect also provides support of
the capacities of working memory and attentional resources. Studies showed that split attention
effects had the tended to increase cognitive load and decrease performance.
Finally, the effect of cognitive load on different types of performance was examined. In
terms of task performance, it appeared that high cognitive load had a negative effect unless the
task was rehearsed to proficiency. In social performance, high cognitive load also had a negative
effect when subjects found the method cognitively taxing. The final study, examining decisions
Effects of Cognitive Load 13
during driving also evidenced that subjects will decrease their cognitive task when imposed with
A review of these studies seems to provide a good amount of evidence for the limitations
of the cognitive system. When a high amount of cognitive resources are necessary for a task, it is
harder to perform. It appears that one method of reducing this need for resources is to present
information tailored to each individual component of working memory. Also, the presentation of
information so that it does not have to be mentally integrated can be beneficial. Another
interesting point found in this research is that task proficiency and levels of expertise seem to
reduce the effects of cognitive load. Finally, it seems to be that none of these studies have found
an effective way to actually measure cognitive load itself. Perhaps this indicates a fault in all of
References
David, P., & Hirshman, E. (1998). Dual-mode presentation and its effect on implicit and explicit
Goolkasian, P. (2000). Pictures, words, and sounds: From which format are we best Able to
Kalyuga, S., Chandler, P., & Sweller, J. (1997). Levels of expertise and user-adapted formats
Modeling: Proceedings of the Sixth International Conference, UM97, Vienna, New York.
Mayer, R., & Moreno, R. (1998). A Split-Attention Effect in Multimedia Learning: Evidence for
90(2), 312-320.
Mayer, R., Moreno, R., Borrie, M. & Vagge, S. (1999). Maximizing Constructivist Learning
Mousavi, S. Y., Low, R., & Sweller, J. (1995). Reducing Cognitive Load by Mixing Auditory
Pontari, B. A., & Schlenker, B.R. (2000). The Influence of Cognitive Load on Self-Presentation:
Velayo, R. S., & Quirk, C. (2000). How do Presentation Modality and Strategy Use Influence
Yeung, A. (1999). Cognitive Load and Learner Expertise: Split-Attention and Redundancy
Effects of Cognitive Load 15