Polite web-based intelligent tutors: Can they improve learning in classrooms?

Abstract

Should an intelligent software tutor be polite, in an effort to motivate and cajole students to learn, or should it use more direct language? If it should be polite, under what conditions? In a series of studies in different contexts (e.g., lab versus classroom) with a variety of students (e.g., low prior knowledge versus high prior knowledge), the politeness effect was investigated in the context of web-based intelligent tutoring systems, software that runs on the Internet and employs artificial intelligence and learning science techniques to help students learn. The goal was to pinpoint the appropriate conditions for having the web-based tutors provide polite feedback and hints (e.g., “Let’s convert the units of the first item”) versus direct feedback and hints (e.g., “Convert the units of the first item now”). In the study presented in this paper, 132 high school students in a classroom setting, grouped as low and high prior knowledge learners according to a pre-intervention knowledge questionnaire, did not benefit more from polite feedback and hints than direct feedback and hints on either an immediate or delayed posttest, both of which contained near transfer and conceptual test items. Of particular interest and contrary to an earlier lab study, low prior knowledge students did not benefit more from using the polite version of a tutor. On the other hand, a politeness effect was observed for the students who made the most errors during the intervention, a different proxy for low prior knowledge, hinting that even in a classroom setting, politeness may be beneficial for more needy students. This article presents and discusses these results, as well as discussing the politeness effect more generally, its theoretical underpinnings, and future directions.

Full text

Page 1

Author's personal copy
Polite web-based intelligent tutors: Can they improve learning in classrooms?
Bruce M. McLarena,*, Krista E. DeLeeuwb, Richard E. Mayerc
aHuman-Computer Interaction Institute, Carnegie Mellon University, Pittsburgh, PA 15213-3891, USA
bKnowledge Media Research Center, Tübingen, Germany
cUniversity of California, Santa Barbara, USA
a r t i c l e i n f o
Article history:
Received 9 August 2010
Received in revised form
27 September 2010
Accepted 28 September 2010
Keywords:
Intelligent tutoring systems
Interactive learning environments
Media in education
Pedagogical issues
a b s t r a c t
Should an intelligent software tutor be polite, in an effort to motivate and cajole students to learn, or
should it use more direct language? If it should be polite, under what conditions? In a series of studies in
different contexts (e.g., lab versus classroom) with a variety of students (e.g., low prior knowledge versus
high prior knowledge), the politeness effect was investigated in the context of web-based intelligent
tutoring systems, software that runs on the Internet and employs artificial intelligence and learning
science techniques to help students learn. The goal was to pinpoint the appropriate conditions for having
the web-based tutors provide polite feedback and hints (e.g., “Let’s convert the units of the first item”)
versus direct feedback and hints (e.g., “Convert the units of the first item now”). In the study presented in
this paper, 132 high school students in a classroom setting, grouped as low and high prior knowledge
learners according to a pre-intervention knowledge questionnaire, did not benefit more from polite
feedback and hints than direct feedback and hints on either an immediate or delayed posttest, both of
which contained near transfer and conceptual test items. Of particular interest and contrary to an earlier
lab study, low prior knowledge students did not benefit more from using the polite version of a tutor. On
the other hand, a politeness effect was observed for the students who made the most errors during the
intervention, a different proxy for low prior knowledge, hinting that even in a classroom setting,
politeness may be beneficial for more needy students. This article presents and discusses these results, as
well as discussing the politeness effect more generally, its theoretical underpinnings, and future
directions.
? 2010 Elsevier Ltd. All rights reserved.
1. Introduction
The abundance of learning materials and programs available on the web today raises an important question: How can we use this new
technology to improve learning? Our research is motivated by the idea that it is important not only to provide easy access to learning
technology, as the web surely does, but also to investigate in a scientific manner ways to make that technology more beneficial to learning.
The field of intelligent tutoring systems (ITS), computer-based learning systems developed with artificial intelligence techniques (VanLehn,
2006), has been providing tutors on the web for some time now (cf. Aleven, McLaren, & Sewall, 2009; Alpert, Singley, & Fairweather,1999;
Beal, Walles, Arroyo, & Woolf, 2007; Melis et al., 2001), and, at the same time, ITS researchers have used an evidence-based approach to
demonstrate impressive improvements in student learning in a range of domains and with different techniques (cf. McLaren, Lim, &
Koedinger, 2008; Mostow & Beck, 2007; Rickel & Johnson, 1999; VanLehn et al., 2005). ITS research builds on the long and substantial
research on instructional feedback, which has demonstrated both significant learning benefits and failures of feedback (Hattie & Gan, 2011;
Kluger & DeNisi, 1996; Shute, 2008). The learning benefits of ITSs have also been traced to specific instructional design principles, such as
minimizing cognitive load and using immediate feedback (Koedinger & Corbett, 2006; Shute,2008). In short, researchon intelligenttutoring
systems has focused on determining what tutors should say to students (i.e., communication content) as well as when theyshould say it (i.e.,
communication pacing).
In contrast, what about the way in which feedback is presented to students and, as a consequence, how (and whether) students perceive
software tutors as learning partners? Much less research has been done on how best to incorporate social cues, such as polite wording, which
* Corresponding author. Tel.: þ1 412 268 8278; fax: þ1 412 268 1266.
E-mail addresses: bmclaren@cs.cmu.edu (B.M. McLaren), k.deleeuw@iwmkmrc.de (K.E. DeLeeuw), mayer@psych.ucsb.edu (R.E. Mayer).
Contents lists available at ScienceDirect
Computers & Education
journal homepage: www.elsevier.com/locate/compedu
0360-1315/$ – see front matter ? 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compedu.2010.09.019
Computers & Education 56 (2011) 574–584

Page 2

Author's personal copy
may be an essential element in student–tutor interactions. Our hypothesis is that intelligent tutors should not only exhibit cognitive intel-
ligencedknowing what to say and when to say itdbut also social intelligencedknowing how to say it. The politeness principle – the idea that
people learn more deeply when instructional support is presented in polite style – is the particular focus of our investigation into social
intelligence. A basis for the politeness principle is suggested, first, by the politeness theory of Brown and Levinson (1987). In their studies of
politeness across cultures, they identified the keyattributes of positive face, likingoradmiringof anotherand/orcooperative and suggestive of
a common goal (i.e., supporting another’s self-esteem), and negative face, being respectful of another’s right to make his or her own decisions
(supporting another’s freedom to act). A second underpinning of our work is the media equation theory of Reeves and Nass (1996) – the idea
that people “respond socially and naturally to media” (Reeves & Nass,1996, p.1), thereby reacting to and treating a computer as a real person.
Finally, the social agencytheoryof Mayeret al. (Mayer, 2005a, 2009) suggeststhat instructional messages – including feedback andhints from
web-based tutors – should be presented in a way that provides social cues, such as polite wording. In this way, the learner may accept the
tutor as a conversational partner, resulting in increased effort to learn the given material, which in turn leads to a better learning outcome.
In the study reported in this paper, we had three goals in testing the politeness principle. First, we sought to carefullyand empirically test
the principle, following from the approach and findings of prior studies, some of which demonstrated clear learning effects (McLaren,
DeLeeuw, & Mayer, in press; Wang et al., 2008), some which demonstrated weaker effects (Wang & Johnson, 2008) and some which did
not show any effects (McLaren, Lim, Yaron, & Koedinger, 2007). Second, we intended to test the specific conditions that influence the
effectiveness of the principle. In particular, we were interested inwhether the effect could survive transition from a lab environment, where
it previously had been successfully employed to promote learning (McLaren et al., in press; Wang et al., 2008; Wang & Johnson, 2008), to
a classroom situation where it had not been successful previously (McLaren et al., 2007). We also intended to test whether low prior
knowledge learners benefit more from polite feedback than high prior knowledge learners, as they had in an earlier study we conducted
(McLaren et al., in press). Finally, a key goal was to test the politeness principle applied in the context of the worldwide web, rather than
within a standard computer software environment. Because the web is omnipresent in today’s classrooms, determining whether, and under
what conditions, politeness makes a difference to learning with web-based materials could open the door to much wider, and scientifically
verified, use of the politeness principle within intelligent tutoring systems.
In the study described in this paper, 132 U.S. high school students who were enrolled in a chemistry class learned from a web-based
intelligent tutor designed to teach stoichiometry, a subtopic of chemistry. Following the learning phase, students completed an immediate
posttest and a delayed posttest one week later, with both near transfer and conceptual questions on both tests. Some students worked with
aversionof the Stoichiometry Tutorthat provided polite problem statements, feedback, and hints, while others worked with aversionof the
tutorthat provided directproblem statements, feedback, and hints. The polite versions of problem statements, feedback, and hints are based
on the face-saving techniques of Brown and Levinson (1987), described briefly above and in more detail in the next section. The study took
3 to 5 hours for students to complete, with most of the work occurring in the classroom (The only work that occurred outside of the
classroom was by students who missed class.). The software tutor was developed using authoring software specifically designed to build
web-based intelligent tutors (Aleven, McLaren, Sewall, & Koedinger, 2009).
In this paper, we briefly review the theoretical underpinnings of the politeness principle, describe in more detail the empirical studies
that have been done to date with the principle, describe the current study and its results, and discuss what we have learned about the
politeness principle thus far.
2. How theory informs our operationalization of the politeness principle
According to Brown and Levinson (1987), politeness has a role in all human culture. Face is the public presentation of such politeness.
Positive face is “the want of every member that his wants be desirable to at least some others,” (Brown & Levinson,1987, p. 62) or, in other
words, the desire of people to be appreciated and approved of by others. Positive face is also characterized by the desire to work cooper-
atively with others. Negative face, on the other hand, is “the want of every ‘competent adult member’ that his wants be unimpeded by
others,” (Brown & Levinson,1987, p. 62), that is, a person’s freedomof autonomy. Positive face refers to a person’s self esteem, while negative
face refers to a person’s freedom to act. According to Brown and Levinson, these two components of human desire are fundamental to any
social interaction, and politeness is an attempt by the participants in those interactions to maintain each other’s faces. Brown and Levinson
have documented the similar ways inwhich people from diverse cultures use politeness tactics for making requests that minimize threats to
both positive and negative face.
Our goal in the present research was to phrase the problem statements, hints, and feedback of our intelligent tutor so as not to threaten
the positive or negative face of the student user, in a manner suggested bythework of Brownand Levinson. Forexample, weworded hints of
our polite tutor to reduce the threat to positive face by presenting cooperative statements (e.g., “Let’s calculate the result now” or “Let’s
convert the units of the first item”) and reduce threats to negative face by allowing freedom of action (e.g., “Do you want to put 1 mol in the
numerator?” or “You may want to convert the units of the first term”). Contrast this with the direct version of our tutor that uses more direct
wording that threatens positive face by being less cooperative (e.g., “The goal here is to calculate the result”) and threatens negative face
through authoritative statements (such as “Put 1 mol in the numerator”). In other words, the theoretical motivation for using polite tutors in
the present study was to prime the learner for social cooperation which will lead, we hypothesize, to deeper learning.
A second theoretical foundation of our work is the media equation theory of Reeves and Nass (1996), which proposes that people can be
induced to relate to a computer as if it is a real person. When social cues are present, Reeves and Nass claim, people easily accept a computer
as a social partner, a claim that is supported by their study in which people learning with computers treated the computers politely, similar
to the way human beings would be treated. Both Reeves and Nass (1996) and Nass and Brave (2005) provide evidence that people need only
a minimal amount of priming to accept a computer as a social partner. Grice (1975) argues that the speaker in a conversation agrees to
generate a message that is intended to make sense to the listener (i.e., the speaker agrees to be clear, relevant, concise, and truthful), and the
listener agrees to exert effort to make sense of the message. When a learner accepts a computer tutor as a social partner, the learner views
a tutor’s message as part of a conversation, which is subject to the rules of conversationdincluding a commitment by the learner to try to
make sense of what the tutoris saying.If the learner has made that commitment, he or she should also process the information moredeeply,
B.M. McLaren et al. / Computers & Education 56 (2011) 574–584
575

Page 3

Author's personal copy
leading to better learning. In the current study, we seek to use polite feedback and hints as away to encourage students to viewaweb-based
tutor as a social partner and to prime the conversational stance.
Finally, Mayer et al. (Mayer, 2005a, 2009) have proposed social agency theory as an extension of the cognitive theory of multimedia
learning. Social agency theory is based on the idea that instructional messages – including feedback and hints from web-based tutor – may
be presented in a way that does or does not involve social cues (e.g., does or does not use polite conversational style). When a tutor’s
message contains appropriate social cues, such as polite wording, the learner accepts the tutor as a conversational partner, which results in
increased effort to engage in cognitive processing aimed at making sense of the tutor’s message, thereby creating a higher quality learning
outcome. When the tutor’s message does not contain social cues, such as when the message is directly or authoritatively expressed, the
learner is less likely to accept the tutor as a conversational partner, and therefore the learner is less likely to work hard to make sense of the
tutor’s message, resulting in a lower quality learning outcome. The cognitive processes that lead to better learning are spelled out in the
cognitive theory of multimedia learning (Mayer, 2005a, 2009), and include selecting relevant information, mentally organizing it into
a coherent structure, and integrating it with other knowledge.
Based on the sum of these theories, we hypothesize that students who work with polite tutors will learn more, both about the specific
material presented and the concepts underlying that material, thanwill students who learnwith direct tutors. We also hypothesize that this
politeness effect will be strongest for students who have low rather than high prior knowledge. Students with high prior knowledge are
more likely to engage in deep cognitive processing during learning without social inducement, and, in fact, polite inducement could even be
distracting to these students. However, students with low prior knowledge are more likely to need some inducement, such as being drawn
into a social interaction with their tutor, to engage in deeper processing.
3. Relevant previous work with intelligent tutors and politeness
Politeness has been an area of interest within the field of intelligent tutoring systems since at least the mid 1990s. The earliest work may
be that of Person, Kreuz, Zwaan, and Graesser (1995) inwhich they found evidence that politeness strategies are commonly used in one-on-
one tutoring interactions between humans, although not always effectively. Their study of human tutoring dialogues suggests that
politeness could, under some circumstances and in different domains, inhibit effective tutoring. They also found that different steps in the
tutoring process appear to be more or less likely to benefit from politeness. However, these early findings were observational, not subjected
to large-scale empirical study, and also not tested with software tutors.
More recently, Mayer, Wang, Johnson and colleagues have performed a series of studies to investigate whether politeness in educational
software, in the form of positive and negative face-saving feedback, can better support learners (Mayer, Johnson, Shaw, & Sandhu, 2006;
Wang et al., 2008). They implemented positive and negative face feedback in the context of the Virtual Factory Teaching System (VFTS),
a factory modeling and simulation tutor. In a polite version of VFTS they have developed, constructions such as, “You could press the ENTER
key” and “Let’s click the ENTER button” were used. Such statements are arguably good for positive face, as they are likely to be perceived as
cooperative and suggestive of a common goal, as well as for negative face, as they are also likely to be perceived as respectful of the student’s
right to make his or her own decisions. In the direct version of VFTS, the tutor used more imperative, direct feedback such as, “Press the
ENTER key” and “The system is asking you to click the ENTER button.” These statements are arguably not supportive of positive face, as they
do not suggest cooperation, or of negative face, as they are likely to be perceived as limiting the student’s freedom.
In a preliminarystudy (Mayeret al., 2006) students wereasked toevaluate the threat tonegative and positive face of a tutor’s statements.
The results indicated that learners are verysensitive to politeness in tutorial feedback, and that learners with less computerexperience react
to the level of politeness in language more than experienced computer users. In the follow-up study run by Wang et al. (2008) in which 37
students were randomly assigned either to a polite tutor group or to a direct tutor group, students who used the polite tutor scored
significantly higher on a posttest. Importantly, the politeness effect was obtained for non-engineering students but not for engineering
students, thus pointing to the notion that students with less prior knowledge of a domain are more susceptible to positive learning effects
from the politeness principle. Wang and Johnson (2008) also observed a significant learning effect due to politeness with a foreign language
tutoring system for a particular type of question, utterance formation questions, in which participants answer a question by recording their
own speech. The participants in this study were paid volunteers, largelywithout significant prior language training, and thus, as in the Wang
et al. (2008) study, low prior knowledge learners. In our own earlier lab study with university students (predominantly psychology majors)
using the Stoichiometry Tutor (McLaren et al., in press), there was a significant pattern in which students with low prior knowledge of
chemistry performed better on subsequent problem-solving tests if they learned from the polite tutor rather than the direct tutor (d ¼ 0.64
on an immediate test, d ¼ 0.50 on a delayed test), whereas students with high prior knowledge showed the reverse trend (d ¼ ?0.58 for an
immediate test; d ¼ ?0.21 for a delayed test). In sum, these studies suggest, first, that the level of politeness in the system feedback of
a tutoring system can make a difference in motivating students and promoting better learning and, second, that the effect tends to be
stronger, and perhaps only useful for students who have less knowledge in the particular domain of interest.
Yet not all research supports the idea that politeness in intelligent tutoring systems will benefit learning. For instance, the Wang and
Johnson (2008) study discussed above, while obtaining a politeness effect for utterance formation questions, did not uncover a polite-
ness learning effect on subjects’ overall score, or an effect on multiple-choice questions or questions involving matching of phrases in Iraqi
Arabic (the language being tutored) to translations in English. In an earlier classroom study involving the Stoichiometry Tutor McLaren et al.
(2007), as opposed tothe lab study of McLaren et al. (inpress), did not find a politeness effect for high school students in a classroom setting.
Although the polite group performed slightly better than the direct group on a posttest, the difference was not statistically significant. Why
did the earlier experiment notobtain a politeness effect whereas both our own, as well as other experiments did? One potentially important
difference is that the learners in this experiment were students taking a college prep chemistry course with a strong and recent background
in chemistry, whereas the learners who produced a politeness effect in the previous experiments were largely unfamiliar with the material.
Another key difference is the setting: The McLaren et al. (2007) study was conducted in classrooms, whereas all of the other politeness
studies, including our own (McLaren et al., in press), were lab-based studies. In the present experiment, we further explore both the use of
the Stoichiometry Tutor in a classroom setting and the idea that low prior knowledge students are most likely to display a politeness effect.
B.M. McLaren et al. / Computers & Education 56 (2011) 574–584
576

Page 4

Author's personal copy
4. Method
4.1. Design
We conducted a study with a 2x2 between-subjects factorial design. One factor was politeness, with one level polite instruction (i.e., use
of the Polite Stoichiometry Tutor) and the other level direct instruction (i.e., use of the Direct Stoichiometry Tutor). The other factor was
modality of feedback, with one level being text only and the other level audio-and-text.1
4.2. Participants and conditions
One hundred and thirty-two (132) high school students (72 female and 60 male) in five chemistry classes in three suburban high schools
in two states (Massachusetts and New Jersey) participated. There were sixteen additional students who at least partially participated, but
scored 0 or very nearly 0 on one or both of the two posttests; these students were eliminated from consideration. We also eliminated two
students because the number of total messages they saw, hints requested plus error messages, was more than 3 standard deviations above
the mean, indicating they were “gaming” the system (i.e., trying to simply get the answers in the last hint, see Baker et al., 2008; Wood &
Wood, 1999). The study materials were used as a replacement for normal lectures and class work on stoichiometry within the five high
school classes, and the three participating teachers were given and used the immediate and delayed posttests as class grades for their
students.
Students were randomly assigned to one of the four conditions of the 2 ? 2 design (polite/text, polite/audio, direct/text, and direct/
audio). The number of students that were assigned to each condition, and in total to the polite and direction conditions, is shown in Table 1.
4.3. Materials and procedure
Table 2 provides an outline of the materials and procedure used in the study. The left-hand column indicates the materials presented to
and completedbystudents in the politeconditions (i.e., both polite/text and polite/audio) and the right-hand column indicates the materials
presented to and completed by students in the direct conditions (both direct/text and direct/audio). The rows between the thick horizontal
lines in the middle of the table represent the intervention materials; these varied bycondition, as further indicated by the highlighting of the
direct materials on the right. All materials were completed online, within a web browser, in the order shown in Table 2. Students used
school-provided computers and headphones, so students could privately listen to the videos and hear audio feedback and hints. All
participants were given user-IDs and passwords that allowed them to logoff and log back on whenever desired.
Because of the usual difficulties in using and tightly controlling classroom time, the study materials of Table 2 were tackled mostly, but
not exclusively, during teacher-monitored classroom time. In a few cases, due to absences or insufficient classroom time, the consent form,
questionnaires, videos, and intervention materials were completed (or viewed) outside of regular classroom time, at school or home. These
materials took students between 60 and 120 min to complete. Two posttests were administered, one immediate and one delayed by one
week. Each posttest took between 45 and 60 min to complete. All of the students took the posttests in class.
The pre-questionnaire contained basic demographic questions, as well as questions about the student’s background in and under-
standing of chemistry. The chemistry questions are shown in Table 3. Answers to these questions were used to separate students into low
and high prior knowledge groups in subsequent analyses. For the first question a score of 1 (“Far below average”) to 5 (“Highly above
average”) was given to each student. For the second question a score between 0 and 10 was given to each student, based on whether they
selected “None of the above are true” (0) or the number of items selected, between 1 and 10. The scores of the two questions were added
together and the mean of all the students’ scores was calculated. All students who scored below the calculated mean of 11.3 were classified
as “low prior knowledge learners,” while all students above the mean were classified as “high prior knowledge learners.” Note that we did
not administer a pretest due to the possibility of “testing effects” (Johnson & Mayer, 2009; Roediger & Karpicke, 2006), the well-studied
phenomenon inwhich student performance improves from taking a test. Since this raises the possibility of washing out learning effects due
to the intervention, we elected to instead ask students to self assess their knowledge.
After completing the pre-questionnaire, the students worked on the intervention materials, including videos and the Stoichiometry
Tutor, that were specific to their condition, as shown in Table 2. The videos were interspersed throughout the materials and were short
(1–4 min), presenting various background materials on chemistry concepts relevant to stoichiometry (e.g., molecular weight, dimensional
analysis), as well as tips on how to solve stoichiometry problems (e.g., problem solving strategy). As indicated in Table 2, the language used
in the videos (i.e., the narration) was specific to condition – polite language was used in the polite condition videos (e.g., “Let’s discuss how
Table 1
Distribution of subjects across conditions.
Polite Instruction Direct Instruction
Text Only
Audio and Text
Polite/text (33)
Polite/audio (31)
Polite (64)
Direct/text (32)
Direct/audio (36)
Direct (68)
1While we originally intended to investigate the learning benefit of providing feedback and hints with a human voice and printed text versus text alone, we later decided
that our design had a redundancy effect (Mayer, 2005b) that would make positive learning effects unlikely. Thus, we focused only on politeness in this study and only the
polite vs. direct aspect of the study will be discussed in the remainder of this paper.
B.M. McLaren et al. / Computers & Education 56 (2011) 574–584
577

Page 5

Author's personal copy
composition stoichiometry allows us to determine the make-up of a molecule .”) and direct language was used in the direct condition
videos (e.g., “Composition stoichiometry refers to the make-up of a molecule.”).
Fig. 1 shows the Stoichiometry Tutor, developed using the Cognitive Tutor Authoring Tools (Aleven, McLaren, & Sewall, 2009; Aleven et
al., 2009), as well as an example of both its polite (Fig. 1a) and direct (Fig. 1b) version. Solving a stoichiometry problem involves under-
standing basic chemistry concepts, such as unit conversions (e.g.,1 g ¼ 1000 mg, as in Fig.1) and the mole, and applying those concepts in
solving simple algebraic chemistry equations. To solve problems, the student must fill in the terms of an equation, cancel numerators and
denominators appropriately, self-explain the reason for each term of the equation (e.g., see the entry of “Given Value” below the first term of
Fig. 1a and b), and calculate and fill in a final result. The student can request hints by selecting the “Hint” button in the upper right-hand
cornerof the interface. If thenumber typed (or unitorsubstance or reason selected) iscorrect,thetyped (or selected)information appears in
Table 2
Materials used and design of the study.
Polite Direct
mr o F t ne s noC d e s a b - beW m r o F t ne s noC de s ab - beW
e r i ann o i t s euQ - e r P e r i ann o i t s euQ - e r P
Five videos, in polite language:
•
Introduction to the Stoichiometry Study
•
Overview of the Tutor and Interface
•
Stoichiometry Problem Solving Strategy
•
Dimensional Analysis
•
Significant Figures
Five videos, in direct language:
•
Introduction to the Stoichiometry Study
•
Overview of the Tutor and Interface
•
Stoichiometry Problem Solving Strategy
•
Dimensional Analysis
•
Significant Figures
Polite Stoichiometry Tutor - Problem # 1 Direct Stoichiometry Tutor - Problem # 1
Polite Stoichiometry Tutor - Problem # 2 Direct Stoichiometry Tutor - Problem # 2
Video: Molecular Weight (in polite language) Video: Molecular Weight (in direct language)
Polite Stoichiometry Tutor - Problem # 3 Direct Stoichiometry Tutor - Problem # 3
Polite Stoichiometry Tutor - Problem # 4 Direct Stoichiometry Tutor - Problem # 4
Video: Composition Stoichiometry (in polite
language)
Video: Composition Stoichiometry (in direct
language)
Polite Stoichiometry Tutor - Problem # 5 Direct Stoichiometry Tutor - Problem # 5
Polite Stoichiometry Tutor - Problem # 6 Direct Stoichiometry Tutor - Problem # 6
Video: Molarity (in polite language) Video: Molarity (in direct language)
Polite Stoichiometry Tutor - Problem # 7 Direct Stoichiometry Tutor - Problem # 7
Polite Stoichiometry Tutor - Problem # 8 Direct Stoichiometry Tutor - Problem # 8
Polite Stoichiometry Tutor - Problem # 9 Direct Stoichiometry Tutor - Problem # 9
Polite Stoichiometry Tutor - Problem # 10 Direct Stoichiometry Tutor - Problem # 10
e r i ann o i t s euQ - t s oP e r i anno i t s euQ - t s oP
Video: Introduction to Post Test Video: Introduction to Post Test
Immediate Posttest:
8 Problems (4 near transfer; 4 conceptual)
Immediate Posttest:
8 Problems (4 near transfer; 4 conceptual)
Delayed Posttest: (One week later)
8 Problems (4 near transfer; 4 conceptual)
Delayed Posttest: (One week later)
8 Problems (4 near transfer; 4 conceptual)
B.M. McLaren et al. / Computers & Education 56 (2011) 574–584
578