Effect of Speaker Accent on Students’ Learning and Speaker-Rating: A Replication

| Volume 2 Index February 2009 |

Full articles

| PDF | SWF |

Title

Effect of Speaker Accent on Students’ Learning and Speaker-Rating: A Replication

 Authors

Paul Kelvin Ong, Vernese Liao, and Rosemarie Alimon

De La Salle University, Manila, 

Philippin

 

Abstract

This study seeks to present the significant effects of speaker accent on students’ deep learning, as measured by retention and transfer tests, and its effects on the students’ rating of the speaker. A narration of how lightning storms develop was conducted among 150 high school students from a private school in Manila, Philippines. Participants were randomly assigned to three conditions: Filipino English–accent, American English–accent, and Korean English–accent. The MANOVA was used to test the effect of the 3 accents on retention and transfer tests, and speaker ratings. The findings showed that students’ performances on retention tests are better when the narration is a foreign-accented voice (American English-accent) (M = 1.48), while students’ performances on transfer tests (M = 4.6) and speaker-rating (M = 6.06) are better and more positive when the narration is in a standard-accented voice. The results were consistent with the Social Agency Theory and the Cognitive Load Theory.

 

Introduction

Listening to the teacher is important in the learning process. When teachers start speaking in class, students will pay attention to learn, and respond to them in various ways. Likewise, how students respond to the information they receive is dependent on the teacher. 

A speaker, such as a teacher, can be effective depending on how he delivers his speech, or lessons. Hand gestures, voice modulation, accent, and pitch are some criteria in evaluating the delivery of a speaker. However, there are other considerations in discerning the effectivity of a speaker—the audience. In a classroom setting, the audience, also known as the students, play an essential role in telling whether a teacher is a good instructor. In a study by Mayer, Sobko, & Mautone (2003), the researchers theorized that a speaker’s accent affects the students’ deep learning (i.e., a speaker who has a standard-accent positively affects the students’ deep learning). However, Davis, Johnsrude, Hervais-Adelman, Taylor, & McGettigan (2005) stated that “Humans are able to understand speech in a variety of situations that dramatically affect the sounds that reach their ears. They can understand talkers with quite different (foreign or regional) accents.”  In experiments conducted by Bent & Bradlow, 2003; Clarke & Garrett, 2004 & Weill, 2001, it was shown “that effective perception of speech in an unfamiliar accent can require several minutes or more of exposure to allow full comprehension” (in Davis, Johnsrude, Hervais-Adelman, Taylor, & McGettigan, 2005). This means that an accent which is unusual may create more cognitive load to the students. Cognitive load refers to the amount of information that is acquired by a learner. 

Recent studies show that more and more students rely on multimedia learning.  Developers of computer-based multimedia instruction are working to create learning experiences in which the learner will accept the computer as a social partner, mainly through the use of on-screen agents that talk to the learner (Cassell, Sullivan, Prevost, & Churchill, 2000 in Mayer, Sobko, & Mautone, 2003).  This is usually experienced in websites, compacted disc encyclopedias, and the like.  However, we are not provided with much research on the role of voice in supporting learning from multimedia lessons.  There are instances wherein the accent of the speaker will also vary to the standard-accent of the listeners.  In this study, researchers consider the underlying role of the speaker’s voice or accent in the multimedia lesson, specifically voice narration, in the influence it can make to the outcome of the students learning.

Mayer, Sobko, and Mautone (2003) examined about the role of speaker’s voice in students’ deep learning through multimedia lessons. They used speakers’ accent as an agent in knowing the effect of a speaker’s voice to learning, which were measured through retention and transfer test. It also explored the effect of accent on the speaker-ratings of the students. The study, however, failed to explore on the role of a country’s second language affects the deep learning of students. The current study seeks to investigate on the effect of a speaker’s accent, particularly, English across three different accents (Filipino English, American English, Korean English) on the learning and speaker-rating of the students having English as their second language.

 

Cognitive Effort Theory

“Effort can be physical and cognitive” (Kanfer, 1992 in Yeo, & Neal, 2008). The current study is centered on cognitive effort. The Cognitive Effort Theory states that challenging situations, where students are pushed to more cognitive work, exert more effort in doing such tasks (Mayer, Sobko, & Mautone, 2003). When students are given a task which they are not familiar with, meaning, it has not been incorporated in their schemas, they will tend to work harder to accomplish the task. In relation to the current study, the theory assumes that students who listen to a speaker with a foreign-accent will perform better on retention and transfer tests since they are presented with a more difficult task (Mayer, Sobko, & Mautone, 2003). In the same way, when the tasks presented to the students are not abundant in resources, they will perform higher if more effort is given to the task (Kanfer, 1987; Norman & Bobrow, 1975 in Yeo, & Neal, 2008).

 

Cognitive Load Theory

The Cognitive Load Theory posits that there is a very limited short-term memory (Miller, 1956 in Tuovinen, & Sweller, 1999), with an infinite long-term memory (Simon & Gilmartin, 1973 in Tuovinen, & Sweller, 1999), having large amounts of schemas (Chi, Glaser, & Rees, 1982, in Tuovinen, & Sweller, 1999), which differ in the level of automaticity (Kotovsky, Hayes, & Simon, 1985, in Tuovinen, & Sweller, 1999). In relation to the current study, the theory assumes that students receiving information from a speaker with a standard accent must utilize fewer cognitive works to process information than receiving data from a human voice with a foreign accent (Mayer, Sobko, & Mautone, 2003). If students have acquired an accent in their schemas which is similar to the standard accent, they will process the given information easier, thus creating lesser cognitive load. Gaining information from a speaker with a foreign accent will in turn produce more cognitive load since the foreign accent has not been incorporated in the students’ schemas. When this happens, students allocate more time in understanding words separately, rather than processing the relationships of the words in the sentence as a whole. For example, students who receive information from a speaker in the standard accent will treat the statement “Water then evaporates” as a single thought because the accent used is automatic in their schemas. Students who receive information from a speaker in the foreign accent will treat each word in the sentence individually, which creates higher cognitive load.  “In line with cognitive load theory (Chandler & Sweller, 1991; Sweller, 1988), Mayer and his colleagues also assume that performance during knowledge acquisition is dependent on the cognitive resources available for information processing” (Brünken, Steinbacher, Plass, & Leutner, 2002). That is, students who acquire knowledge from a standard-accented voice will perform better in problem-solving transfer tests as supported by the Social Agency Theory (Mayer, Sobko, & Mautone, 2003).

 

Social Agency Theory

The Social Agency Theory posits that when the learners’ social conversation schema is activated, they are more likely to imagine as if they are having a conversation with another individual.  Therefore, they are somehow engaging in the social rules of human-to-human contact (Mayer, Sobko, & Mautone, 2003) In return, learners try harder to make sense of what the speaker is saying by engaging in deep cognitive processing.  This was referred to as cooperation principle since the four conversational maxims are being experienced, wherein listeners assume that the speaker is trying to make sense by being informative, accurate, relevant, and concise (Grice, 1975 in Mayer, Sobko, & Mautone, 2003). In the current study, it is hypothesized that students who receive a narration which comes from a standard-accented voice (Filipino English-accent) will perceive the speaker possessing the four conversational maxims. 

Learners engage deeper in explanation of the text when they think as if the author is speaking to them (Beck, McKeown, Sandora, Kucan, & Worthy, 1996; Schraw & Brunning, 1996 in Mayer, Sobko, & Mautone, 2003).  In the same way, students who experience conversational classroom settings have better performances than lecturing (Cohen, Kulik, & Kulik, 1982; Graesser, Person, & Magliano, 1995 in Mayer, Sobko, & Mautone, 2003).  In addition, a lecture based type of teaching can be seen as a learner receiving weak social cues such as listening to a foreign-accented voice. The learner would most likely interpret this merely as a case of information delivery.  Thus, the cognitive processing would only be seeking for gaining information rather than going further by understanding (Mayer, Sobko, & Mautone, 2003). In the present study, the researchers find support in this theory as it is assumed that students who listen to a narration by a standard-accented voice (Filipino English-accent) will create stronger social cues for them, thus, gaining better performances in tests of transfer.   

This study seeks to present the significant effects of the accent of the narrative voice in the students’ deep learning. Assuming that the participants would understand their standard accent voice rather than a foreign accent voice, the researchers will study which of the three groups would perform better in the retention and transfer tests, and would rate the speaker positively.  Students’ performance on tests of transfer is better when the narration in a narrated animation comes from a standard-accented voice rather than a foreign-accented voice.  Moreover, students will produce higher ratings of the speaker on socially desirable characteristics when the speaker uses a standard-accented voice rather than with a foreign accented voice. 

In the study, the three theories were integrated through the manipulation of the independent variables such that the social agency theory, cognitive load theory and the cognitive effort theory were tested all at the same time using different tests. Social agency theory can be connected with the cognitive load theory such that both theories suggest that learners learn better when they are exposed to a standard accented voice as compared to a foreign accented voice; however, the social agency theory also takes into consideration speaker rating which the cognitive effort theory does not test. The cognitive effort theory, which posits that learners learn more when they are exposed to a challenging situation, is also tested in contrast with the social agency theory and cognitive load theory.

 

Method

Participants and Design

The researchers randomly selected students from a private school in Quezon City. The sample included 150 high school students, ranging from second to fourth year. All students are Filipino citizens who reside in the Philippines, with a mean age of 15.46. Fifty participants were assigned in the American English–accent (foreign–accent) group (n males = 57.50%; n females = 42.50%), 50 participants for the Korean English–accent (foreign–accent) group (n males = 32.26%; n females = 67.74%), and 50 participants were assigned in the Filipino English–accent (standard–accent) group (n males = 41.18%; n females = 58.82%). The design used in the study was a simple-randomized group design for three independent variables. The independent variable was kind of English accent, with three levels, Filipino English–accent, American English–accent, and Korean English–accent. The dependent variables were students’ scores in the retention test, four transfer tests, as well as the speaker-rating survey.

 

Materials

The materials consisted of a participant questionnaire, a retention test, four transfer tests, and a speaker-rating survey. The participant questionnaire asked about demographic information such as their name, age, gender, year level and nationality. 

The current study only exposed participants to a recorded human voice with foreign-accent (American and Korean English) and standard-accent (Filipino English) narrating about lightning; since the main focus was about the social cues in multimedia learning. 

The participants received a narration which explained how lighting forms. The narration was spoken by a female which had a Filipino English – accent (standard-accent), a female with an American English–accent, or a female with a Korean English–accent. All narrations were recorded thru a laptop microphone. The narration was adapted from a study by Mayer, Sobko, and Mautone (2003) as shown in Appendix A. The apparatus used was a compact disc player with loud speakers for each of the three groups.

Retention Test. The retention test required the participants to write down an explanation of how lightning works (see Appendix B). This test is considered to be a retention test because participants are tasked to write information that was presented through the narration. The test also measured what the students recalled from the narration. 

Transfer Tests. The four transfer tests were distributed separately after giving the retention test. The first transfer test asked what they could do to decrease the intensity of lightning (Appendix C).  The second transfer test brought them in a situation wherein suppose they see clouds in the sky but there is no lightning and why not (see Appendix D).  For the third transfer test, they were asked what the air temperature has to do with lightning (see Appendix E).  Finally, for the fourth transfer test participants were asked what causes lightning (see Appendix F). The transfer tests require participants to choose and use certain, specific, and relevant information from the presentation and must find ways on how they can connect it to their answers. In addition, the four transfer tests measured how well the students comprehended the narration. Hence, transfer tests go beyond recall of information, but it actually uses the recall of information as a component in order to come up with sound solutions for the questions being asked. Take for instance the first transfer test question, wherein the participants must recall a specific aspect of lightning explanation, particularly that negative particles from the cloud meet positive charges from the ground. From there, he must be able to use that recalled explanation to arrive at inferred ideas in order to answer the question by suggesting that positive particles should be removed from the ground. Although the idea was not directly presented in the narration, it can be implied, and therefore, represents a form of transfer. 

Speaker-rating Survey. The speaker-rating survey was a 15-item instrument cited in the study by Mayer, Sobko, and Mautone (2003), which they adapted from Zahn and Hopper’s (1985) Speech Evaluation Instrument. At the top of the survey questionnaires were instructions asking the participants to encircle a number from 1 to 8 indicating how the speaker (multimedia human voice) sounded along each of the 15 dimensions. For each dimension, the numbers 1 to 8 were indicated along a line with one adjective at the top of 1 and an opposite adjective at the top of 8. The 15 adjective pairs were the following: illiterate-literate, unkind-kind, passive-active, unintelligent-intelligent, cold-warm, shy-talkative, uneducated-educated, unfriendly-friendly, aggressive-unaggressive, not fluent-fluent, unpleasant-pleasant, unsure-confident, inexperienced-experienced, unlikable-likable, and lazy-energetic. By averaging the scores, an overall speaker-rating was achieved, with 1 indicating the most negative and 8 indicating the most positive. According to Mayer, Sobko, and Mautone (2003), they adapted the Speech Evaluation Instrument by Zahn and Hopper (1985) because it was effective in detecting the social characteristics attributed to speakers.

 

Procedure

Participants were randomly selected by the school such that 50 students were assigned to each of the three groups. The experiment was conducted in air-conditioned rooms to minimize the noise and distractions which might greatly affect the participants’ listening to the narration. First, the experimenters introduced themselves and told the students that they will be participating in an exercise which involves listening. After, the participants were asked to complete the participant questionnaire. Then, the experimenter told the participants that they shall be receiving an explanation on how lightning storms develop and reminded the participants to maintain silence during the process of listening to the narration because questions shall be asked to them afterwards. The recorded narration was then played through a compact disc player with loud speakers. On the basis of random assignment, participants in the standard-accent group (Filipino English–accent) received a standard-accent version (Filipino English–accent) of the narration, and the participants in the two foreign-accent groups (American English–accent and Korean English–accent) received foreign-accent versions (American English–accent and Korean English–accent) of the narration respectively. When the narration ended, the participants in each group were tasked to answer a retention test and asked students to keep writing until told to stop. After four minutes, the retention test was collected and the first transfer test sheet was distributed. After two and a half minutes, the first transfer test sheet was collected and the next transfer sheet was distributed and so on until all four transfer sheets had been completed. Then, the experimenters explained to the participants the instructions in completing the speaker-rating survey; it was clarified that the recorded voice was the one to be rated. After the participants had completed the speaker rating survey, it was collected. Finally, participants were thanked through incentives (a notebook and a pencil for each participant) and debriefed about the study.

 

Data Analysis

The retention test was scored by counting the number of ideas that a participant was able to write (maximum of eight points) disregarding specific wording. The ideas were as follows: (a) air rises, (b) water condenses, (c) water and crystals fall, (d) wind is dragged downward, (e) negative charges fall to the bottom of the cloud, (f) the leaders meet, (g) negative charges rush down, and (h) positive charges rush up. 

Scores from the transfer tests were obtained by tallying the number of acceptable answers given across all four questions (with a maximum of two points per test; and eight points for all the transfer tests). Some acceptable answers for the first transfer test about decreasing the intensity of lightning were as follows: (a) removing positively charged particles from the ground and/or (b) placing positively charged particles near the clouds. For the second transfer test about seeing clouds but no lightning, some acceptable answers were: (a) the top of the cloud might not be above the freezing level, and/or (b) there are no negative particles in the cloud. Some acceptable answers for the third transfer test regarding air temperature and lightning included: (a) air must be cooler than the surface of the earth and/ or (b) the top of the cloud must be cooler than the bottom. For the fourth transfer test about the causes of lightning, some acceptable answers included: (a) there must be a difference of electrical charge within the cloud and/ or (b) between the cloud and the ground.          

The researchers used Multivariate Analysis of Variance (MANOVA) to analyze the data. Using the scores of students belonging in the three different groups, values for descriptive statistics was obtained as well as the univariate results for each dependent variable, multivariate test of significance and the Scheffé test.

 

Results

The means and standard deviations on the retention test, transfer tests and speaker-rating measures are indicated in Table 1 (See Appendix G).

Effect of Accented voice on Retention 

There was a significant difference in the means of the participants who were exposed to a standard-accented voice and foreign-accented voice on retention F (2, 147) = 33.50, MS error = 0.8981, p = 0.000. However, the mean of the group exposed to standard-accented voice (Filipino English-accent) in retention (M=1.26) was lower compared to the mean of the foreign-accented group specifically the American English-accent (M= 1.48). The standard deviations on the retention test reveal that the group who listened to a standard-accented English was more dispersed with SD = 1.42 as compared to the other two groups, the American English-accent group SD = 0.79, and the Korean English-accent group with SD = 0.20.

Effect of Accented voice on Transfer 

Conforming with the social agency theory and cognitive load theory, it was found that students performed better on problem solving transfer when the voice was from a human speaking voice with a standard accent (Filipino English-accent) rather than a human speaking voice with a foreign-accent (American English-accent and Korean English-accent). Participants who received the narration from a standard-accented voice scored significantly higher on problem-solving transfer than the participants who received a foreign-accented voice (American English-accent and Korean English-accent) F (2, 147) = 70.99, MS error = 3.30, p = 0.000. From the means of the transfer test, it can be directly observed that the group who received a standard-accent scored higher; (M = 4.6) as compared to other two groups exposed to foreign-accented voice, the American English-accent (M = 1.8) and the Korean English-accent (M = 0.34). Standard deviations on transfer reveal that the group who listened to a standard-accented voice was more dispersed with SD = 2.39 as compared to the other two groups, the American English-accent with SD = 1.83, and the Korean English-accent group with SD= 0.92. Based on the result, it can be implied that students in the standard-accented group exerted more effort in making relevant learning outcomes than did the students in the foreign-accented groups.

Effect of Accented voice on Social Rating of the Speaker

 In line with the social agency theory, participants who received a standard-accented human voice (Filipino English-accent) rated the speaker higher as compared to the two other groups of participants who received a foreign-accented voice  (American English-accent and Korean English-accent) F (2, 147) = 85.68, MS error = 1.884, p = 0.000. From the means of the speaker-rating surveys, it can be seen that participants in the standard-accent group (Filipino English-accent) rated the speaker higher with M = 6.06, than the other two foreign-accent groups, the American English-accent (M = 5.31) and the Korean English-accent (M = 1.55). Standard deviations on transfer reveal that the group who listened to a standard-accented voice was more dispersed with SD = 1.58 as compared to the other two groups with the American English-accent group SD = 1.55, and Korean English-accent group with SD = 0.86.

 

Discussion

Based on the findings of the experiment, it was found that students performed better on problem solving transfer when the voice in the narration came from a human speaking voice with a standard-accent rather than a foreign-accent voice. Filipinos are accustomed to learning through listening to a Native American accent since they speak ESL or English as a Second Language which explains why they scored high on retention.  This is highly inconsistent with the Cognitive Effort Theory which states that students exert more cognitive effort when a particular task is challenging, thus, achieving better performances. The researchers consider the possibility that both Social Agency Theory and Cognitive Load Theory might be accepted in the study. The Social Agency Theory and Cognitive Load Theory explain how a person’s schema is essential in a cognitive process. Regarding the Social Agency Theory, it states that social cues in multimedia narrations activate the schema of individuals, and therefore, they imagine that they are having a conversation with another human. In the present study, what activated the schema of the participants is the narration of a standard-accented voice. As for the Cognitive Load Theory, it posits that when students listen to an accent that is found in their schemas, they will process the narration easier, which creates lesser cognitive load.  

The Social Agency Theory is further supported by the results from the speaker-rating survey, wherein the participants rated the speaker having a standard-accent higher on social dimensions than the speaker who had a foreign-accent. Therefore, voice is a factor that can be considered in creating a sense of social presence in which learners would interpret a narrator from a media as a social partner. In addition, the cooperation principle by Grice (1975) which stated that when students are engaged in deep processing, they perceive the speaker as trying to make sense by being informative, accurate, relevant, and concise. This can explain why the students who listened to the standard-accented voice rated the speaker higher as compared to the other groups.

The Cognitive Load Theory is also supported through the results of the transfer tests of the students. Students who were exposed in a standard-accented voice scored higher than those who were exposed in a foreign-accented voice. Therefore, foreign-accented voice is seen as a negative factor which causes lower transfer scores of students because it creates more cognitive load. Another reason for the low scores of students in the transfer test was that students in the foreign-accented group devoted more cognitive capacity in deciphering the incoming narration, which resulted to less cognitive capacity left to make connections among pieces of information that is essential for meaningful learning. 

Regarding the results obtained in the retention test, the American English-accent group (foreign-accent) scored higher than the standard-accented group (Filipino English-accent). This implies that the American English-accent group was able to devote cognitive capacity to listening, encoding, and remembering the main facts in the narration. In addition, deeper cognitive processing shall be needed for the transfer test, so the researchers are most interested in transfer; because it is through the transfer tests that meaningful learning can be measured. Another explanation regarding this matter is from Davis, Johnsrude, Hervais-Adelman, Taylor, and McGettigan (2005) which posits that humans can understand words even if a speaker has an accent which is foreign to them. 

As for the results regarding the two foreign-accents (American English and Korean English), it suggests that an American English-accent voice would result in better learning of students as compared to a Korean English–accent voice.  English is considered to be the second language in several countries including the Philippines.  Moreover, learners will have difficulty understanding as well as processing the narration from a Korean-accented voice since this language is not very familiar in their schemas.

In the Philippine setting, it would be essential for Filipino schools with a majority of Filipino students (i.e. students who grew up in a home using Filipino as language of speaking) would benefit more in their learning if they would hire Filipino teachers in their school.  Moreover, the results obtained from the experiment imply that a teacher with an American English-accent or Filipino English-accent will not be a problem to students since English is the second language of the Philippines.

 

References

Brünken, R., Steinbacher S., Plass, J. L., & Leutner, D. (2002). Assessment of cognitive load in multimedia learning using dual-task methodology. Experimental Psychology, 49(2), 109-119. 

Davis, M. H., Johnsrude, I. S., Hervais-Adelman, A., Taylor, K., & McGettigan, C. (2005). Lexical information drives perceptual learning of distorted speech: Evidence from the comprehension of noise-vocoded sentences. Journal of Experimental Psychology, 134(2), 

222–241

Hunton, J. E., Hall, T. W., & Price K. H. (1998).  The value of voice in participative decision making.  Journal of Applied Psychology, 83(5), 788-797.

Mayer, R. E., Sobko, K,. & Mautone, P. D. (2203). Social cues in multimedia learning: Role of speaker’s voice. Journal of Educational Psychology, 95(2), 419–425. 

Tuovinen, J. E. & Sweller, J. (1999). A comparison of cognitive load associated with discovery learning and worked examples. Journal of Educational Psychology, 91(2), 334-341.

Yeo, G. & Neal, A. (2008). Subjective cognitive effort: A model of states, traits, and time. Journal of Applied Psychology, 93(3), 617–631.

 

Appendix A

Narration Script for Lightning Lesson

Cool, moist air moves over a warmer surface and becomes heated. Warmed, moist air near the earth’s surface rises rapidly. As the air in this updraft cools, water vapor condenses into water droplets and forms a cloud. The cloud’s top extends above the freezing level, so the upper portion of the cloud is composed of tiny ice crystals. Eventually, the water droplets and ice crystals become too large to be suspended by the updrafts. As the raindrops and ice crystals fall through the cloud, they drag some of the air in the cloud downward, producing downdrafts. When downdrafts strike the ground, they spread out in all directions, producing the gusts of cool wind people feel just before the start of the rain. Within the cloud, the rising air currents cause electrical charges to build. The charge results from the collision of the cloud’s rising water droplets against heavier, falling pieces of ice. The negatively charged particles fall to the bottom of the cloud, and most of the positively charged particles rise to the top. A stepped leader of negative charges moves downward in a series of steps. It nears the ground. A positively charged leader travels up from such objects as trees and buildings. The two leaders generally meet about 165 feet above the ground. The negatively charged particles then rush from the cloud to the ground along the path created by the leaders. It is not very bright. As the leader stroke nears the ground, it induces an opposite charge, so positively charged particles from the ground rush upward along the same path. This upward motion of the current is the return stroke. It produces the bright flash that people notice as a flash of lightning.

 

Appendix B

Retention Test

Please write down an explanation of how lightning works.

PLEASE KEEP WORKING UNTIL YOU ARE TOLD TO STOP

 

Appendix C

Transfer Test 1 Question

What could you do to decrease the intensity of lightning? 

PLEASE KEEP WORKING UNTIL YOU ARE TOLD TO STOP.

 

Appendix D

Transfer Test 2 Question

Suppose you see clouds in the sky but no lightning, why not?

PLEASE KEEP WORKING UNTIL YOU ARE TOLD TO STOP.

 

Appendix E

Transfer Test 3 Question

What does air temperature have to do with lightning?

PLEASE KEEP WORKING UNTIL YOU ARE TOLD TO STOP.

 

Appendix F

Transfer Test 4 Question

What causes lightning?

PLEASE KEEP WORKING UNTIL YOU ARE TOLD TO STOP.

 

Appendix G

Means and Standard Deviations of Retention, Transfer, and Speaker-rating Survey for the three Groups

Category: 2009