Revisiting Modeling Examples: Researching The Influence of Visible and Invisible
Models in a Replication Study
Educational Psychology
200300103
Universiteit Utrecht
03.11.2023
Word count: 1999
Abstract
This study focuses on replicating and examining differences between modeling examples
with and without visible models in a learning context. Drawing from the original study by
Hoogerheide et al. (2014), the experiment investigates test performance, self-efficacy, and
perceived competence among participants exposed to different instructional videos.
Participants, 20 university students, were randomly assigned to view either a video of
a modeling example with a visible model or without one. The pretest assessed prior
knowledge, the learning phase involved video instruction, and the posttest measured
performance, self-efficacy, and competence.
Results showed no significant differences in test performance or self-efficacy before
and after video exposure. However, there was a significant difference in perceived
competence post video viewing, favoring modeling videos with a visible model.
Key words: modeling examples, self efficacy, competence, cognitive theory, worked
example effect, performance
Introduction
Example-based learning has proven to be an efficient and effective strategy for teaching
novice learners to solve new problems and for increasing the students' self-efficacy and
,competence (Gog et al., 2019; Kirschner et al., 2006). Previous research has mainly focused
on the differences between worked and model examples based learning, as is the context of
the original study (Hoogerheide et al., 2014). However, in modelling examples a distinction
is made between model examples with or without a video example. Little research has been
conducted on the differences fostered between these two. In order to gain more knowledge
about these differences and due to time limits, this paper will focus on the partly replication
of modeling examples of the study ‘Comparing the effects of worked examples and modeling
examples on learning’ by Hoogerheide et al., (2014). Replication studies are important for
validating results of previous studies, assessing generalizability of a finding, uncovering
possible biases or methodological flaws of the original experiment and enhancing credibility
(Earp & Trafimow, 2015). This paper will cover important concepts relating to model-based
learning, test performance, self-efficacy and competence.
Model example based learning
Effective learning occurs when students build correct mental representations (Mayer, 2005).
Example based learning helps this process by showing complete step-by-step solutions. This
way, cognitive load is minimized, enabling that all of working memory can be devoted to
learning and the acquisition of cognitive skills is based on the abstraction of underlying
principles of a problem (Gog et al., 2019). It also causes the worked example effect (Gog et
al., 2019).
Model examples are defined by having a domain-specific, competent individual
explain how to complete a task, either on video or live (Huang, 2017). Modelling examples
with a visible model have the model visible in the video, showing how the model adapts
every step of the instruction. Modelling examples without a visible model rely on verbally
explaining the task, with possible computer generated changes of visual aspect to direct
, attention on important concepts to understand underlying structure of the problem
(Hoogerheide et al., 2014). Compared to working models, model based learning adds
additional benefits of having visual aspects and the model’s voice. These factors have proven
to stimulate the learner to link presented content to themselves (Mayer, 2005). Another
benefit is that it enables social cues more than worked examples, which can affect
motivational aspects of learning such as self-efficacy and perceived competence. Having a
visible model increases social clues (Hoogerheide et al., 2014; Mayer, 2005). Lastly, research
of Paivio and Clark’s (2006) dual coding theory enhances the belief on modelling example
benefits. His research suggests that the mind processes information along two different
channels: verbal and visual. He concluded that by combining verbal and visual aids you are
more likely to process and retain knowledge more effectively, minimize cognitive overload,
and maximise memory capacity (Mayer, 2005).
Effects on performance, self efficacy and competence
The different characteristics of with or without a visible model may lead to different effects.
Starting with the cognitive load theory, which argues that keeping mental load on working
memory to a minimum is one of the most important factors for fostering effective learning
(Plass et al., 2010). Split attention effect is closely related to the cognitive load theory, and
refers to using multiple information sources at the same time (e.g., text, sounds and pictures)
(Pouw et al., 2019). Research has been conflicted on whether model examples cause or avoid
split attention. Arguments have been made that model examples could be the cause, because
of the verbal aspect and because having human faces and movement naturally attracts
attention. When that draws attention away from the problem solving instruction, it causes
split attention and negatively impacts working memory (Pouw et al., 2019; Hoogerheide et
al., 2014). However, counter arguments have been made that visible models do not distract
Models in a Replication Study
Educational Psychology
200300103
Universiteit Utrecht
03.11.2023
Word count: 1999
Abstract
This study focuses on replicating and examining differences between modeling examples
with and without visible models in a learning context. Drawing from the original study by
Hoogerheide et al. (2014), the experiment investigates test performance, self-efficacy, and
perceived competence among participants exposed to different instructional videos.
Participants, 20 university students, were randomly assigned to view either a video of
a modeling example with a visible model or without one. The pretest assessed prior
knowledge, the learning phase involved video instruction, and the posttest measured
performance, self-efficacy, and competence.
Results showed no significant differences in test performance or self-efficacy before
and after video exposure. However, there was a significant difference in perceived
competence post video viewing, favoring modeling videos with a visible model.
Key words: modeling examples, self efficacy, competence, cognitive theory, worked
example effect, performance
Introduction
Example-based learning has proven to be an efficient and effective strategy for teaching
novice learners to solve new problems and for increasing the students' self-efficacy and
,competence (Gog et al., 2019; Kirschner et al., 2006). Previous research has mainly focused
on the differences between worked and model examples based learning, as is the context of
the original study (Hoogerheide et al., 2014). However, in modelling examples a distinction
is made between model examples with or without a video example. Little research has been
conducted on the differences fostered between these two. In order to gain more knowledge
about these differences and due to time limits, this paper will focus on the partly replication
of modeling examples of the study ‘Comparing the effects of worked examples and modeling
examples on learning’ by Hoogerheide et al., (2014). Replication studies are important for
validating results of previous studies, assessing generalizability of a finding, uncovering
possible biases or methodological flaws of the original experiment and enhancing credibility
(Earp & Trafimow, 2015). This paper will cover important concepts relating to model-based
learning, test performance, self-efficacy and competence.
Model example based learning
Effective learning occurs when students build correct mental representations (Mayer, 2005).
Example based learning helps this process by showing complete step-by-step solutions. This
way, cognitive load is minimized, enabling that all of working memory can be devoted to
learning and the acquisition of cognitive skills is based on the abstraction of underlying
principles of a problem (Gog et al., 2019). It also causes the worked example effect (Gog et
al., 2019).
Model examples are defined by having a domain-specific, competent individual
explain how to complete a task, either on video or live (Huang, 2017). Modelling examples
with a visible model have the model visible in the video, showing how the model adapts
every step of the instruction. Modelling examples without a visible model rely on verbally
explaining the task, with possible computer generated changes of visual aspect to direct
, attention on important concepts to understand underlying structure of the problem
(Hoogerheide et al., 2014). Compared to working models, model based learning adds
additional benefits of having visual aspects and the model’s voice. These factors have proven
to stimulate the learner to link presented content to themselves (Mayer, 2005). Another
benefit is that it enables social cues more than worked examples, which can affect
motivational aspects of learning such as self-efficacy and perceived competence. Having a
visible model increases social clues (Hoogerheide et al., 2014; Mayer, 2005). Lastly, research
of Paivio and Clark’s (2006) dual coding theory enhances the belief on modelling example
benefits. His research suggests that the mind processes information along two different
channels: verbal and visual. He concluded that by combining verbal and visual aids you are
more likely to process and retain knowledge more effectively, minimize cognitive overload,
and maximise memory capacity (Mayer, 2005).
Effects on performance, self efficacy and competence
The different characteristics of with or without a visible model may lead to different effects.
Starting with the cognitive load theory, which argues that keeping mental load on working
memory to a minimum is one of the most important factors for fostering effective learning
(Plass et al., 2010). Split attention effect is closely related to the cognitive load theory, and
refers to using multiple information sources at the same time (e.g., text, sounds and pictures)
(Pouw et al., 2019). Research has been conflicted on whether model examples cause or avoid
split attention. Arguments have been made that model examples could be the cause, because
of the verbal aspect and because having human faces and movement naturally attracts
attention. When that draws attention away from the problem solving instruction, it causes
split attention and negatively impacts working memory (Pouw et al., 2019; Hoogerheide et
al., 2014). However, counter arguments have been made that visible models do not distract
