Encouraging and facilitating the process of students explaining their own understanding and thinking
in the learning environment
Dasha Bunks
Liberty University
PSYC 775 - Teaching of Psychology
Dr. Winn
2022
Introduction
Self-explaining, or making sense of new information by explaining to oneself, can greatly
enhance learning for students across ages, domains, and instructional types. Over two decades of
research have demonstrated that explaining concepts while learning results in more effective problem-
solving, more robust conceptual understanding, and better monitoring of learning. But how can a
teacher apply these findings to his or her own instruction? In what settings and with what kinds of
students can self-explanations be most effective? This chapter reviews self-explanation research to give
teachers practical advice on how to incorporate self-explanation most effectively into their own
classrooms. What are self-explanations?
Classrooms often present information to be learned through text and/or words from an
instructor. In order to learn, students must take information from various sources and construct
meaning or understanding. Self-explaining, or making sense of new information by explaining to oneself,
helps learners construct new knowledge by elaborating upon presented information, relating them to
existing knowledge, making inferences, and making connections among given information. For example,
in the seminal paper finding the self-explanation effect (Chi, Bassok, Lewis, Reimann & Glaser, 1989),
students learned to solve basic mechanics problems by studying a worked-out example solution from a
college physics textbook. One problem consisted of a written description of a weight hanging from three
strings accompanied by a diagram (taken from Chi et al., 1989).
Examples of self-explanations generated by students working this problem follow in quotations:
The diagram shows an object of weight W hung by strings. Consider the knot at the junction of the three
strings to be the body. “Why should this [the knot] be the body? I thought W was the body” 92 The body
at rest under the action of the three forces shown… “I see. So, the W will be the force and not the body.
OK…” …of the three forces shown in the diagram. “Uh huh, so…so they refer to the point as the body.”
Chi et al. (1989) found that successful students spontaneously generated more self-explanations as
compared to less successful students, that is, they uttered statements that indicated they refined their
understanding of the problem and made connections to relevant principles. Even with no instructor,
coach, tutor, or feedback on the explanation, just the act of engaging with the information through
generating explanations improved learning.
This self-explanation effect, or students learning better when they explain material to
themselves, has replicated across domains such as biology (Chi et al., 1994), computer science (Pirolli &
Recker, 1994; Recker & Pirolli, 1995), probability (Renkl, 1997) electricity and magnetism (Ferguson-
Hessler & de Jong, 1990), and history (Wolfe & Goldman, 2005). More than two decades of research
demonstrate the benefits of self-explanation with wide ranges of ages (e.g. Calin-Jageman & Ratner,
2005; Griffin, Wiley, & Thiede, 2008; Kastens & Liben, 2007; Pillow, Mash, Aloian, & Hill, 2002; Pine &
Messer, 2000) and instructional formats (e.g. Ainsworth & Loizou, 2003; Roscoe & Chi, 2008; Trabasso &
Magliano, 1996).
The self-explanation effect is one of the twentyfive principles of learning from the Association
for Psychological Science (Graesser, Halpern, & Hakel, 2007) and is one of the Institute for Educational
Science's seven recommended learning strategies in the 2007 practice guide (Pashler et al., 2007). Why
are self-explanations helpful? Generally, self-explaining is a constructive activity requiring students to
actively engage in their learning process. Active participation is better than passive participation for
learning (Bransford, Brown, & Cocking, 1999; Chi, 2008).
How does self-explanation, in particular, help people learn? Two mechanisms have the most
empirical support (Chi, 2000). First, self-explanations help learners identify gaps in their understanding
and fill in missing information from the instructional material. For instance, when students are trying to
digest information in a textbook, on the computer, or while solving a problem, there are inherent
omissions in the information. That is, even the best materials cannot possibly contain all the information
that is needed for every learner for the specific topic.
Self-explanations help learners generate inferences to fill in those missing gaps in their own
understanding of the topic. The second way that self-explanations can help learning is by revising what a
student already knows about the topic. Learners can come in with their own ideas, or their own mental
models of a concept. These mental models are typically flawed. When a learner encounters instructional
material that conflicts with their existing mental models, self-explaining helps repair and revise their
understanding.
Supporting Self-Explanation: Description and Discussion
Although good learners typically self-explain more, not all self-explanations are equal. A closer
look at the results from Chi et al. (1989) reveals qualitative differences between good and bad problem
solvers. Good learners tended to spend more time studying examples, give explanations that were more
principle-based, elaborate on the conditions and goals, and monitor comprehension. Poorer learners
made less of these productive types of self-explanations and instead paraphrased and re-read materials.
Additionally, there were no differences in prior domain knowledge or GPA between the good and poor
93 learners (Chi & VanLehn, 1991).
Other studies have replicated the findings that successful learners more often use particular
kinds of self-explanations (Ferguson-Hessler & deJong, 1990; Pirolli & Recker, 1994) independent of
prior domain knowledge. Studies have also reproduced the self-explanation effect with different media,
such as studying with text and diagrams (Ainsworth & Loizou, 2003) or learning from multiple sources on
the web (Goldman, Braash, Wiley, Graesser, & Brodowinska, 2012). Unfortunately, however, studies
also show that only a small percentage of learners naturally generate productive selfexplanations (Renkl,
1997). Building upon the finding that successful learners self-explain, Chi et al. (1994) investigated
whether prompting self-explanations could have similar benefits for learning. That is, if good learners
naturally generate productive self-explanations, could instructional prompts elicit self-explanations from
students to help them learn more deeply?
Chi et al. (1994) asked eighth-grade students to self-explain after reading each sentence of a
passage on the human circulatory system. A comparison group of students read the same passage twice
but were not asked to self-explain. The students who were prompted to self-explain understood more
about the circulatory system than the comparison group. Within the selfexplain group, the students who
generated more self-explanations learned better than students who did not self-explain as often. High
self-explainers had more accurate mental models and understood complexities of the circulatory system
that were not explicit in the text.
A wealth of studies has replicated this finding that explicit prompts for self-explanations
encourages students to generate explanations and improves learning in various settings (e.g. Bielaczyc,
Pirolli & Brown, 1995; McNamara, 2004; Renkl, 1997). For example, Griffin, Wiley, and Thiede (2008)
compared three groups of college students: one group read a complex text once “as though studying for
an exam” (p. 95), the second group read and then re-read the same text as if there was a test, and the
third group read the text but was also instructed to self-explain while reading with a few example
questions such as, “What new information does this paragraph add?” (p. 97).
Students who were instructed to self-explain outperformed either the read-once or read-twice
condition on assessments administered to all groups after the treatment. Thus, explicitly prompting can
help students who may not spontaneously generate self-explanations learn with understanding. Instead
of having a coach or instructional support to prompt self-explanation at specific times, providing
students with training on how to self-explain can help students learn. For instance, Renkl, Stark, Gruber,
and Mandl (1998) investigated the impact of self-explanation training on understanding of compound
interest with a population of vocational students in bank apprenticeships.
A researcher modeled selfexplanation and then provided coaching while learners practiced self-
explaining on a warm-up problem. Learners who received self-explanation training outperformed
learners who were trained to think aloud during instruction. Bielaczyc, Pirolli, and Brown (1995) trained
university students to self-explain while learning about recursive functions in computer science.
Students with self-explanation training explained more and performed significantly better on posttest
problems compared to students that discussed, performed simple recall tasks, and wrote essays.
McNamara (2004) created a selfexplanation reading training program (SERT) that significantly improved
undergraduate students' comprehension of science texts.
Both the Bielaczyc et al. (1995) and McNamara (2004) training generally involved an
introduction of self-explanation, videotaped models of learners self-explaining, and practice self-
explaining with feedback from the trainer before instruction. Eliciting self-explanations may benefit
learning not only in the short term (where assessments are given immediately after instruction) but also
foster transfer (Chi et al., 1994) and retention of concepts. Wong, Lawson, and Keeves (2002) compared
the performance of ninth-grade mathematics students with self-explanation training to a control group
of students using their typical studying methods. Self-explaining students outperformed the control
group on geometry posttests, especially for far-transfer items that required application of the newly
learned material to a novel situation. King (1992) had college students enrolled in a remedial reading
course watch a lecture and take notes. One group of students self-explained lecture content, one group
generated a summary of the content and another group reviewed lecture notes. On posttests
immediately after the session on a retention test one week later, the explainers and summarizers
outperformed students who reviewed lecture notes.
Instructional techniques to encourage self-explanation
Various kinds of instructional techniques can help students self-explain. Many studies have
asked students for verbal self-explanations while studying texts or listening to lectures. Prompting
students for written self-explanations has also been successful (Hausmann & Chi, 2002; Schworm &
Renkl, 2007). Other studies have used computer-based systems to prompt explanations (Chiu & Linn,
2013; Wylie & Chi, in press) or have students select the most appropriate explanation for the given
concept (Aleven & Koedinger, 2002). All of these techniques of eliciting self-explanations have been
shown to benefit learning across many domains and ages.
Computer-based self-explanation programs can greatly increase access to the benefits of self-
explanation. Resource-intensive approaches such as one-to-one human prompting or one instructor
training a whole class at a time require training of human tutors and may or may not provide individual
student attention. Computer-based environments that support self-explanation enable students to get
one-on-one benefit without having a human present. Although students do not tend to spontaneously
type self-explanations as much as they spontaneously verbalize self-explanations (Hausmann & Chi,
2002), studies demonstrate various successful approaches to computer-supported self-explanations
(Recker & Pirolli, 1995; Wylie & Chi, in press).
Some learning environments have students generate and type open explanations to specific
prompts (Chiu & Linn, 2013; Schworm & Renkl, 2006), whereas other studies find benefit from prompts
that involve selecting parts of explanations from drop-down menus (Booth, Lange, Koedinger, &
Newton, 2013). For example, Aleven and Koedinger (2002) investigated the impact of drop-down self-
explanations with an intelligent computer-based mathematics tutoring program in high school geometry
classes. One group of students was prompted to explain each step of the geometry problems, whereas
another group entered answers to each step but was not prompted to self-explain.
The prompted explanation group significantly outperformed the regular problem-solving group
on multiple posttest measures. Similarly, computer-based training can be an effective alternative to
human training for self-explaining. Building upon the success of the human trainers in the SERT program,
McNamara and her colleagues developed a computer-based program called the Interactive Strategy
Training for Active Reading and Thinking (iSTART) to help students understand complex science texts.
iSTART coaches students in the particular successful self-explanation strategies, such as making
inferences, anticipative predictions, or elaborations.
Animated pedagogical agents provide the training by interacting with each other and the
learner. Just like SERT, iSTART involves an introduction, demonstration, and practice of the strategies
through teacher and student agents that give feedback on learners' self-explanations. Studies
demonstrate that iSTART improves reading comprehension scores for students (McNamara, O'Reilly,
Best, & Ozuru, 2006). 95 Optimal conditions for self-explanation Although prompting explanations
generally enhances learning, the actual explanation may be fragmented, partially correct, or even
entirely wrong (Renkl, 2002). Helping students make better selfexplanations and finding optimal
conditions for self-explaining can make learning through selfexplanation more effective. Studies have
explored factors that may contribute to the quality or the impact of self-explanations, including the prior
knowledge of the learner, instructional material and specific types of explanation prompts.
With regard to prior knowledge, self-explanation can help students with a range of expertise and
abilities. Self-explanation has been found to be beneficial for low-knowledge students (O'Reilly, Best, &
McNamara, 2004; Renkl et al., 1998) or students with no prior knowledge of the subject (DeBruin,
Rikers, & Schmidt, 2007). Some researchers suggest that there may be a greater benefit of
selfexplanations with more knowledge to draw upon (Wong, Lawson, & Keeves, 2002). Many studies
find self-explanation beneficial regardless of prior knowledge (Chi et al., 1994; Griffin, Wiley, & Thiede,
2008). The lack of a clear trend in these studies indicates that self-explanations can benefit students
with different abilities in different ways.
Learners may use self-explanations to fill in gaps of understanding or to repair existing mental
models depending on their prior knowledge and the specific instructional context (e.g. Chi, 2000). Other
studies investigated how the format of material to be learned may influence the self-explanation effect.
Findings suggest that if the instructional material contains diagrams, learners tend to generate more
self-explanations (Butcher, 2006). For example, Ainsworth and Loizou (2003) compared learners with a
text passage about the circulatory system to another group that had diagrams with text. Learners with
the diagrams generated significantly more self-explanations and performed significantly better than
those with only text. Learning with multiple diagrams, or multiple forms of media, may benefit from
prompting selfexplanations (Roy & Chi, 2005; Wylie & Chi, in press). Rau, Aleven and Rummel (2009)
found that sixth grade students learned more about fraction concepts using multiple graphical
representations when prompted for explanations that helped students integrate information. In the
Aleven and Koedinger (2002) study, the prompted explanations specifically helped students integrate
verbal and visual information into robust declarative knowledge.
Prompting self-explanations may help students use multiple representations in complementary,
constraining, or constructing roles to fill in missing pieces or revise their understanding (Ainsworth,
1999; Chi, 2000). The correctness and coherence of the instructional material can also impact how
learners self-explain. Multiple studies show that including incorrect solutions or examples in study
materials can be particularly helpful for learning with self-explanation (Booth et al., 2013; Siegler &
Chen, 2008). Durkin and Rittle-Johnson (2012) grouped fourth and fifth grade students learning about
decimals into two groups, one compared correct with incorrect examples and the other only compared
correct examples. Students with both incorrect and correct examples learned more and made more
connections back to concepts in their explanations. Self-explaining incorrect examples provides an
excellent opportunity for students to either recognize gaps in understanding or repair faulty
information, whereas explaining correct examples may be more likely to reaffirm students' existing
understanding.
References
Allen, K. P. (2010). Classroom management, bullying, and teacher practices. The Professional Educator,
34(1), 1-15.
Allen, J., Gregory, A., Mikami, A., Lun, J., Hamre, B., & Pianta, R. (2013). Observations of effective
teacher-student interactions in secondary classrooms: Predicting student achievement with
the Classroom Assessment Scoring System — Secondary. School Psychology Review, 42 (1), 76-
98.
Allensworth, E. M., & Easton, J. Q. (2007). What matters for staying on-track and graduation in Chicago
public high schools. Consortium of Chicago School Research at the University of Chicago.
Retrieved from https://ccsr.uchicago.edu/sites/default/files/publications/07%20What
%20Matters%20Final.pdf
Ang, R. P. (2005). Development and validation of the teacher-student relationship inventory using
exploratory and confirmatory factor analysis. The Journal of Experimental Education, 71(1), 55-
74.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs,
NJ: Prentice Hall.
Baker, J. A. (2006). Contributions of teacher-child relationships to positive school adjustment during
elementary school. Journal of School Psychology, 44, 211-229.
Battistich, V., Schaps, E., & Wilson, N. (2004). Effects of an elementary school intervention on students'
"connectedness" to school and social adjustment during middle school. The Journal of Primary
Prevention, 24(3), 243-262.
Berry, D., & O'Connor, E. (2009). Behavioral risk, teacher-child relationships, and social skill
development across middle childhood: A child-by-environment analysis of change. Journal of
Applied Developmental Psychology, 31(1), 1-14.
Birch, S. H., & Ladd, G. W. (1997). The teacher-child relationship and early school adjustment. Journal of
School Psychology, 55(1), 61-79.
Birch, S. H., & Ladd, G. W. (1998). Children's interpersonal behaviors and the teacher-child relationship.
Developmental Psychology, 34(5), 934-946.
Borman, G. & Overman, L. (2004). Academic resilience in mathematics among poor and minority
students. Elementary School Journal, 104(3), 177-195.
Bowlby, J. (1969). Attachment and loss: Volume 1. Attachment. New York: Basic Books.
Brown, J. L., Jones, S., LaRusso, M. D., & Aber, J. L. (2010). Improving classroom quality: Teacher
influences and experimental impacts of the 4Rs program. Journal of Educational
Psychology,102(1), 153-167.
Bryk, A. S., Sebring, P. B., Allensworth, E., Luppescu, S., & Easton, J. Q. (2010). Organizing schools for
Improvement: Lessons from Chicago. Chicago, IL: The University of Chicago Press.
Cassidy, J. & Asher, S.R. (1992). Loneliness and peer relations in young children. Child Development, 63,
350-365.
Conduct Problems Prevention Research Group. (2010). Effects of a multiyear social-emotional learning
program: The role of student and school characteristics. Journal of Consulting and Clinical
Psychology, 78(2), 156-168.
Croninger, R. G., & Lee, V. E. (2001). Social capital and dropping out of high school: Benefits to at-risk
students of teachers' support and guidance. Teacher College Record, 103(4), 548-581.
Curby, T. W., LoCasale-Crouch, J., Konold, T. R., Pianta, R. C., Howes, C., Burchinal, M., Bryant,
D., ...Barbarin, O. (2009). The relations of observed pre-k classroom quality profiles to
children's achievement and social competence. Early Education and Development, 20(2), 1556-
6935.
Curby, T. W., Rimm-Kaufman, S. E., & Ponitz, C. C. (2009). Teacher-child interactions and children's
achievement trajectories across kindergarten and first grade. Journal of Educational
Psychology,101 (4), 912-925.
Daniels, D. H., & Perry, K. E. (2003). "Learner-centered" according to children. Theory Into Practice,
42(2), 102-108.
Deci, E. L., & Ryan, R. M. (2002). Overview of self-determination theory: An organismic dialectical
perspective. In E. L. Deci & R. M. Ryan (Eds.), Handbook of self-determination research (pp. 3-
33). Rochester, NY: University of Rochester Press.
