Gender and computer-mediated communication: an exploration of elementary students' mathematics and science learning.

Although computer-mediated communication (CMC)--its technology, its contents, and usage patterns--is still in the process of rapid change, the use of CMC such as the Internet as a teaching-learning tool is increasing dramatically. However, very few research studies have examined the gender differences of the characteristic of communication and interaction. This study, therefore, examines gender differences in student communication and interaction in the context of mathematics and science learning using CMC.

In this study, 22 elementary students' interaction and communication patterns using CMC are examined. The results of the data analysis show that, first, it remains clear that gender is still a factor that needs to be considered in the context of mathematics and science learning using CMC. In terms of communication patterns, male students, compared to their female counterparts, are more likely to present their opinions and explanations, but less likely to making specific suggestions; whereas female students tend to ask for a lot of information, but are less likely to provide explanations and/or present their opinions. Particularly, they are less likely to start discussions by "giving explanation." In addition, female students tend to initiate conversations, while male students are more likely to enter the dialogue at later stages and respond to previous discussions.

**********

Because of the belief that there should be equity for females and males in mathematics and science education, the gender issue is an active line of research in the US and other countries (King & McLeod, 1999) and has been studied intensely for about three decades. These research results promote numerous efforts to improve educational practices for all students in schools. Although schools are making progress toward equitable treatment of male and female students, concerns remain (American Association of University Women [AAUW], 1999). In mathematics and science, male students still outscore female students and the gender gap increases with grade level on nationally representative tests as well as on the Third International Mathematics and Science Study (TIMSS)-- "an achievement test given to half a million fourth, eighth, and twelfth-grade students in 41 nations in 1995-96" (AAUW, 1999). Findings from previous studies (Fennema & Sherman, 1978; Li, 1999) that dealt with interactions in the mathematics and scie nce classroom demonstrated a typical pattern: that is, male students tend to interact more than female students. Other studies show that investment in a collaborative learning environment dramatically improve students', especially female students', mathematics learning. For instance, Scott and Heller (1991) found that collaborative learning helped females achieve more, feel more successful, and value mathematics and science more.

COMPUTER-MEDIATED COMMUNICATION

Although computer-mediated communication (CMC)--its technology, its contents, and usage patterns--is still in the process of rapid change, the use of CMC, such as the Internet as a teaching-learning tool is increasing dramatically.

CMC, first developed in 1970 by Murray Turoff(Harasim, 1990), is an effective electronic means of connecting learners without time and location constraints using computers (Machtmes & Asher, 2000). The educational advantages of this medium have been documented. It has been claimed that CMC is a "device which provides a framework for group collaboration from a distance and which, in pedagogy, can enhance collaborative learning" (Henri, 1996). Wells states that CMC can:

...provide students with opportunities for convenient course-related or social interaction with peers...enable collaborative group work by distance students ... facilitate interaction with an instructor ... decrease turnaround time for instructor feedback ... allow students access to on-line resources, e.g., databases, library catalogues, and course registration ... and enable students to upload and download assignments and take on-line quizzes and tests ... (Wells, 1992, p. 2)

Kaye summarized the advantages of using CMC in education in terms of:

- the convenience of an asynchronous communication mode, which liberates users from both time and space constraints;

- its value as a medium of written communication, within a system in which students are graded essentially on the quality of their written work;

- the enhanced levels of interactivity between and amongst students, tutors, course developers, and other members of a widely dispersed learning community;

- the reduction of the isolation felt by many distance learners, and the potential of CMC for collaborative learning. (Kaye, 1992, p. 228)

It is noted that the CMC not only gives students powerful learning tools, it also affects their thinking and behavior. However, we still know little about "how computer technology is affecting the educational gap between girls and boys" (AAUW, 1999). Previous research (Lipponen & Hakkarainen, 1997) suggest that computer-supported collaborative learning is particularly motivating for those students who have traditionally been having a lot of difficulties with adapting to traditional school learning. In addition, using CMC to provide a collaborative learning environment tends to help decrease the gender gap concerning student interaction. Although the previously cited research findings suggest that the collaborative learning environment provided by CMC may help narrow the gender gap in student mathematics and science learning, which includes student learning behaviors and outcomes, computers are beginning to create a new gender gap as indicated by the American Association of University Women (1999).

Therefore, it is essential to the field of mathematics and science education to assess the role of new technology in promoting equity and collaboration. In cognitive and social psychology, there has already been much research about how male and female students communicate and interact, but the context has usually been the traditional face-to-face setting with its paralinguistic cues to facilitate communication. What we do not yet know is how male and female students differ in terms of their communication and interaction in the CMC context. Questions asked to identify the patterns of participant communication and interaction in CMC settings tackle one of the most important issues in research on CMC (Aviv & Golan, 1998; Harris & Jones, 2000) and they should be "posed--and solved--well before the task of setting up an operational telelearning system within the educational framework can be tackled" (p. 203).

A networked community is defined as a group of users who interact and communicate using networked computers. Harris and Jones (2000) suggested that research of social interaction and communication among members of a networked community is one of three primary approaches to research on CMC. Findings of research studies of participants' communication and interaction patterns will help in the design of educational technologies and learning environments to ensure the inclusion of elements that are effective for different types of learners. A better understanding of male and female student mathematics and science learning patterns in CMC settings will allow for increased usefulness of the CMC tools and enhanced mathematics and science teaching and learning in the future. However, because the educational use of CMC is a relatively new venture for most teachers and students (Heaviside et al., 1997), there is limited research on any kind of educational telecommunication in K-12 classrooms. Of that limited offering, v ery little has examined the nature of social interaction among participants in a networked community (Harris & Jones, 2000), and even less the gender differences of the characteristic of communication and interaction. This study, therefore, examines gender differences in student communication and interaction in the context of mathematics and science learning using CMC.

LANGUAGE FUNCTIONS

While there are a wide variety of useful methods to investigate communication, one of the specific focuses of this study is on the in-depth analysis of the language functions in message content that is very crucial to student learning process (Henri, 1992).

Language functions refer to what the learner can do with language, such as giving and receiving information, asking for clarification, and expressing agreement or disagreement (McDonell, 1992). Saville-Troike (1996) indicated that the functions of language generally provide the primary dimension for characterizing and organizing communicative processes. They further elaborated that without understanding why a language is being used as it is, one cannot understand the meaning of its use in the context of social interaction.

Group collaboration focuses on negotiation of meaning in which various language functions appear. Among them are the functions of asking information, giving information, and explaining. Giving and requesting information are basic functions used during group collaboration. People make suggestions, express opinions, agree and disagree about ideas. They negotiate meanings and try to arrive at a common understanding by discussing and debating. Although several researchers have subsumed the categories of language functions under other headings and elaborated them in diverse ways, their purposes are similar (Finocchiaro & Brumfit, 1986).

The classification of language functions used in the present study is the result of incorporating the work of Olsen (1992) and Olsen and Kagan (1992). The combination of their work provides a more complete category of language functions. Among the various language functions being categorized, the five language functions of "asking information," "giving information," "making suggestion," "presenting opinion," and "expressing disbeliefs" are chosen for this study. The reason for choosing these particular language functions is that they are viewed as the basic functions of group work and are essential to its use in mathematics and science contexts (Kessler, Quinn, & Fathman, 1992). The operational rules used in coding the language functions (see appendix) are adapted from Cookson (1995) and Olsen (1992). The integration of the two categorizing systems allows a better fit with the needs of the study and is more appropriate to the context of the study--namely, mathematics and science learning uses CMC.

Interaction

According to Harasim (1990), collaborative learning theory highlights group interaction. Interaction is a very important indicator of collaborative learning and interaction is often cited as the major learning factor in collaborative learning (O'Donnel & Dansereau, 1992). Analysis of interaction allows us to describe the actual structuring of CMC content. Moreover "the analysis of interactivity can lead to an evaluation of the levels of collaboration at work among learners, of their active participation in the accumulation of knowledge, and of their skills in structuring the information presented onscreen" (Henri, 1992, p. 128). The importance of the concept of interaction in CMC made it imperative to provide some in-depth analysis. For the purpose of the present study, interaction is analyzed in terms of the length and generation of messages. Definitions of these related key terms are needed first to help set the context.

Message tree is one of the key terms used in this study. A message tree is defined as a group of messages that interact with one another either explicitly or implicitly. In this study, the following operational definition is used to classify messages created in a CMC setting.

1. Interactive messages: these answer or interpret a previous statement-they refer to the theme of the teleconference and are also connected, explicitly or implicitly, to one or more other messages.

2. Non-interactive or independent messages: their content relates to the theme of the teleconference, but is not connected to other messages of the teleconferences. (Henri, 1995, p. 152)

A message tree takes a hierarchical structure as shown in Figure 1. The circles denote to individual messages in a tree, while the arrows connecting different circles represent interactive relationships between messages. A message is interacted with another message if it is an interactive message as defined by Henri (1995). If a message interacts with a previous message, the second message is "referenced" to the first message indicated by an arrow. The arrows pointing the messages from one to another in Figure 1 is the reference of this type.

The idea of using a "tree" metaphor to represent this structure is two fold. One is that the structure takes a "fruity tree" shape. The other is that the author believes that the "tree" analogy properly represents the interaction processes in the CMC context and is particularly suitable for this study. Generally speaking, messages are the outcomes of people's collaboration. People read, think, understand, summarize, elaborate, explain, question, and discuss, which all are part of their interaction processes (branches), while the messages (fruit) hung on the "trees" stem from their interaction (branches). Moreover, language functions associated with messages are the "seeds" contained in "fruits." People plant "seeds," grow "trees," and finally they get "fruits."

Another key term used in this study is generation. Every message in a tree is assigned a generation value according to its position in the tree. If a message does not refer to any other messages, it is the message at generation 1. Otherwise, if a message that refers to an earlier message, it is at generation of that earlier message plus one (Figure 2). For the remaining of this article, the term generation is used only to describe this specific value of a message. Messages at generation 1 are defined as "initial messages."

Purpose and Research Questions

The main purpose of this study is to examine the gender differences in student communication patterns using CMC. To examine this, male and female student use of the five language functions in their mathematics and science learning using the CMC are compared. Furthermore, male and female student use of the five language functions in the initial messages are explored. A secondary interest in this study is to investigate gender differences in student interaction in their mathematics and science learning using CMC. To explore this gender difference, the length and generation of messages created by male and female students are compared.

The following questions guide this study:

1. Do male and female students differ in their use of the five language functions in their messages?

2. Do male and female students differ in their use of the five language functions in their initial messages?

3. Do male and female student messages differ in terms of the length?

4. Do male and female student messages differ in terms of the generation value?

Context of the Study

The focus of the study is to exam gender differences in elementary student mathematics and science learning in the context of CMC. The actual system used is the Knowledge Forum. The medium is an example of a more generic type commonly referred to as CMC media.

Knowledge Forum is essentially a students and their teacher generated communal database. According to Scardamalia and Bereiter (1996), the key features of the system include tools for the creation and storage of messages, and means to direct students toward collaboration. Taking together these elements, Knowledge Forum provides an environment for more purposeful processing of information. In Knowledge Forum, students and their teachers can read and create text or graphic messages to share their ideas. They can set up views to represent various facet of their collaborative work. This communal database enables every participant access the data and work to build higher levels of understanding. The system provides (a) means that encourage theory formulation and sustained inquiry and (b) database search mechanisms that assist students generating collaborative communities in which they work to advance the ideas in them (Scardamalia & Bereiter 1996).

More importantly, Knowledge Forum facilitates knowledge generation and construction. In Knowledge Forum, according to Scardamalia and Bereiter (1996), knowledge construction rather than knowledge retrieval or navigation is the key issue. In such a dynamic setting, students are constantly engaged in higher-level cognitive processes:

* explanation, problem finding, problem solving, and so forth;

* in-depth understanding of informative texts;

* active, intentional learning; and

* literacy and literacy-related problems (Scardamalia & Bereiter, 1996)

METHODS

Subjects and Procedures

The subjects for the study are selected from students who participated in the "Knowledge Forum Knowledge Building Community Project" in 1999-2000. It is a sixth-grade classroom in an inner city elementary school in Toronto, Ontario. There are 22 students in the class with 11 female students and 11 male students. The data--the entire corpus of electronic messages from their mathematics and science course--are used for the analysis. These data are already collected and provided by the Knowledge Forum Knowledge Building Community Project team in electronic form.

The transcripts are coded for language functions (e.g., asking information, giving information, making suggestions, presenting opinions, and expressing disbelief) and generation. The unit of analysis is the individual message. Each message is coded as either an interactive massage or an independent message following Henri's (1995) definition as presented earlier. It is then assigned a generation value. For each of the five language functions, every message that contains at least one sentence using the language function is coded as that language function message. For instances, if a message has one or more sentences used "giving explanation," this message is coded as a "giving explanation" message. Because it is highly possible that a message contains sentences using more than one language function, a message then can be coded as more than one specific language function message (e.g. a message can be both a "giving explanation" message and a "presenting opinion" message). The length of a message is measured by number of T-units in the message. A T-unit is defined as one:

...main clause plus what ever subordinate clauses happen to be attached or embedded within it. Thus "Mary hit John" is one T-unit, and "Mary hit John, but she is my best fried" is a two T-unit. (Crookes, 1990, p. 185)

Here, a clause is:

...one subject or one set of coordinate subject with one finite verb or one finite set of coordinated verbs. Thus "I went home" is one clause, and so is "Jim and I went home and rode our bikes." (Hunt, 1966, p. 735)

RESULTS

Of the 506 student messages collected, 381 messages are analyzed. These messages are selected because they are considered to be dialogical: that is, the messages are interactive or showed an intention to be interactive. These dialogical messages include those "interactive" messages and "independent" messages as defined by Henri (1995, P. 152). Messages that are not heading toward a dialogue (here called nondialogical messages) are excluded. These nondialogical messages constituted messages of merely "quotations" or "rise-above," offered without comment. Here, "rise-above" is a standard scaffold provided by the software.

Gender Difference Related to Communication

To examine gender difference with respect to student communication in the context of CMC, both descriptive and inferential statistics are conducted to detect patterns. Together these results answer research questions 1 and 2.

First, proportions of the male and female student messages using each of the five language functions are compared. Percentages of the use of the five language functions in all messages by male and female students are represented in Figure 3. Perhaps the most striking patterns are the gender differences in the use of "giving explanation" and "presenting opinion." Compared to the female student messages, male student messages contain proportionally more messages using "giving explanation" and "presenting opinion." However, female student messages include proportionally more messages using "asking information" than male student messages. For male student messages, "presenting opinion" is the language function used most often among the five language functions examined through the current study, followed by "asking information." In addition, nearly a third of the male student messages are "giving explanation" messages. For female students, "asking information" is the language function that had been used most frequ ently among the five language functions, while "giving explanation" messages only constituted about one fifth of their messages. The proportion of "asking information" messages drastically exceeds that of the other four language function messages.

Percentages of the use of the five language functions in initial messages by male and female students are represented in Figure 4. The patterns of student use of the five language functions in initial messages are parallel to the patterns identified in all messages.

Second, descriptive statistics for the five language functions used in all student messages and in student initial messages are examined. These data are presented in Tables 1 and 2 respectively.

Next, inferential statistics are conducted to explore gender difference with respect to student use of the five language functions. Because the sample is not normally distributed, nonparametric rather than parametric statistics are used. Mann-Whitney U tests are administrated to compare male and female student messages, alpha is 0.05.

* Language Functions in All Messages. The differences between male and female student use of the five language functions in all messages are examined. These results are presented in Table 3. On average, female student messages contain significantly more "asking information" messages than do male student messages, p = 0.024 (Table 3). The mean number of "asking information" is 0.40 for male student messages, and 0.52 for female student messages (Table 1). However, analyses of student use of four language functions ("giving explanation," "presenting opinion," "making suggestion," and "expressing disbelief') by gender show no statistically significant difference (Table 3).

* Language Functions in Initial Messages. Patterns identified from previous experiences in face-to-face discourse indicated that the initial talk often largely affects the development of the whole conversations. To study the possibility that the initiative talk might have similar effects in CMC settings, the differences between male and female student use of the five language functions in initial messages are examined using Mann-Whitney U tests. These results are presented in Table 4. On average, female initial messages have significantly fewer "giving explanation" initial messages than do male initial messages, p = 0.03 (Table 4). The mean number of "asking information" messages is 0.22 for male student messages, and 0.13 for female students (Table 2). However, analyses of four language function messages ("asking information," "presenting opinion," "making suggestion," and "expressing disbelief") by gender show no statistically significant difference (Table 4).

Gender Difference Related to Interaction

The examination of gender differences with respect to interaction is conducted from the following two perspectives: length and generation of messages. Length of a message is measured in terms of the number of T-units in the message. Descriptive statistics are presented in Table 5.

T-tests are administrated to examine the differences between male and female student messages in terms of the number of T-units and the generation value of a message. These results are presented in Table 6. On average, the number of T-units of female student messages is significantly greater than that of male student, p <0.001 (Table 6). The mean number of T-units is 2.96 for male student messages, and 3.65 for female student messages. That is, compared to male student messages, female student messages tended to be longer. In addition, the generation value for male student messages is significantly greater than that of female student messages. The mean number of generations is 2.32 for male student messages, and 2.03 for female student messages. This suggested that, compared to male students, female students tend to initiate discussions.

CONCLUSIONS

Several issues and concerns emerge related to student communication and interaction pattern using CMC. First, it remains clear that gender is still a factor that needs to be considered in the context of mathematics and science learning using CMC. In this study, the number of messages (249, 65.4%) generated by male students is nearly double the number of messages (132, 34.6%) generated by female students. This is consistent with previous studies (Li, 1999; Herring, 1993, 1994) showing that female students are less interactive than their male counterparts in the context of CMC. The relative low participation of female students in the electronic discussion may be due to the fact that female students generally have negative feelings about using computers (Shashaani, 1997).

Patterns of male and female student use of language are also explored. Among the five language functions studied, the language function observed most frequently is the "presenting opinion" for male students. "Presenting opinion" and "giving explanation" messages constitute the majority (75.1%) of male student messages. The language function that is least commonly used by male students is "making suggestions." This shows that male students are more likely to present their opinions and explanations rather than making specific suggestions. This is parallel to what Harris and Jones (2000) found when examining school students and teachers electronic communication. In Harris and Jones' study, participants reported information, especially personal and general information and ideas, opinions, most often.

For female students, different patterns emerged. "Asking information" is the language function that is used most frequently. Proportionally, the use of this language function drastically exceeds the other four language functions. Compared to their male peers, female students tend to ask for a lot of information, but are less likely to provide explanations and/or present their opinions. Particularly, they are less likely to start discussions with "giving explanation." This is consistent with previous findings (Yates, 1997; Herring, 1993) that females are more likely to post queries than are males. Comparing the female student initial messages to their general messages, similar language patterns appeared.

Another interesting finding of this study is that male students, on average, created messages that are much shorter than female students. This also concurs with earlier research findings (Barrett & Lally, 1998; Stewart, et al, 1998; Herring, 1992, 1993) that females post shorter messages than males. No definite conclusions can be drawn from this study. Further research investigating this phenomenon is needed.

Probably the most interesting finding of this study is that female students, compared to their male counterparts, used significantly more "asking information" in their messages. In addition, male student messages have significantly higher generation value than female student messages. This shows that female students tend to initiate conversations, often by asking questions, while male students are more likely to enter the dialogues at later stages and to respond to previous discussions. This demonstrates a pattern of female communication in online environments--starting with questions and then waiting until others (most likely male students) jump in to provide information and/or offer their opinions. This fits with previous research by Tannen (1994, 1996), who looked at adults' communication patterns in face-to-face settings. Tannen suggested that females like to interact and learn by asking questions. They gain knowledge from the answers, and their increased knowledge motivates them to explore new ideas and ask more questions.

However, findings of the current study and Tannen's work conflict with Barret and Lally's (1998) study of college students' (11 female and 5 male) interaction in an online learning environment. The result of their study demonstrated that females, compared to their male counterparts, tended to respond to previous messages. Since there are a number of differences among these studies, including the participants' age and the learning environments (i.e., face-to-face vs. CMC), more research is needed to determine the cause of these differences.

Regardless of the pattern, the results of the current study suggest that there are gender differences between male and female behaviors in the online context. These gender differences further support the idea that online environments appear to replicate many of the power structures (e.g., gender) found in face-to-face interactions (Yates, 1997). As researchers have indicated (Yates, 1997; Herring, 1993), although the nonverbal cues disappeared in CMC, gender is reflected in the language used. These gendered identities and practices, as mirrored in face-to-face communications, cause the male dominance of discussions.

Many educators and researchers have high hopes for CMC, believing that it provides more equal access to information and communication, and ultimately leads to a greater equity (Graddol & Swann, 1989; Landow, 1992). This study of gender differences in the context of CMC shows that male and female students demonstrated different interaction and communication patterns. These male and female behavior differences, as reflected in the current study as well as others (Herring, 1993; Yates, 1997), further support the idea that gender identities are reflected in online communicative styles (Herring, 1994). Consequently, "gender differences, along with their social consequences, are likely to persist on computer-mediated networks" (Herring, 1994, p. 5). This suggests that simply moving learning environments from face-to-face to CMC contexts cannot automatically achieve greater equity.

However, this should not diminish the importance of CMC. Rather, it calls for further studies of what causes gender differences in CMC environments. These types of studies may shed light on where gender differences come from in learning environments of all sorts, including face-to-face settings and CMC contexts. It may also provide additional ideas for appropriate teaching and learning strategies for this new learning environment. Given the research at this point, it is not viable to provide ways of most effective use of CMC in classrooms. We need additional experimentations and studies of ways in which to break down those gender roles in this new environment so we can use that more successfully. This study also suggests that we need to consider different types of learners in designing educational technology and learning environment. For the increased usefulness of CMC and enhanced mathematics and science learning in the future, a better understanding of male and female student learning patterns in a CMC envi ronment is necessary.

Figure 3

Percentage of messages used specific language functions by gender


    male  female

AI  40.2  52.3
GE  32.5  23.5
PO  42.6  34.1
MS  8     11.4
ED  11.6  11.4

Note: Table made from bar graph

Figure 4

Language functions used in first messages


    male  female

AI  27.7  34.1
GE  22.1  12.9
PO  30.5  24
MS   4.8   4.5
ED   6.8   5.3

Note: Table made from bar graph

Table 1

Mean Number of Students' Use of Language Functions

Language functions    Male    Female  Total

Asking information    249 *   132     381
                        0.40    0.52    0.44
                        0.49    0.50    0.50
Giving explanation    249     132     381
                        0.32    0.23    0.30
                        0.46    0.43    0.46
Presenting opinion    249     132     381
                        0.43    0.34    0.40
                        0.49    0.11    0.49
Making suggestion     249     132     381
                        0.08    0.11    0.09
                        0.27    0.31    0.29
Expressing disbelief  249     132     381
                        0.12    0.11    0.12
                        0.32    0.31    0.32

* For each language function the first row of numbers is the number of
messages in the specified group; the second row of numbers is the mean
for the specified group; and the third row of numbers is the standard
deviation of the specified group.

Table 2

Mean Number of Students' Use of Language Functions in Initial Messages

Language functions    Male    Female  Total

Asking information    249 *   132     381
                        0.27    0.34    0.30
                        0.45    0.48    0.46
Giving explanation    249     132     381
                        0.22    0.13    0.20
                        0.42    0.34    0.39
Presenting opinion    249     132     381
                        0.31    0.24    0.28
                        0.46    0.43    0.45
Making suggestion     249     132     381
                        0.05    0.05    0.05
                        0.21    0.20    0.21
Expressing disbelief  249     132     381
                        0.07    0.05    0.06
                        0.25    0.22    0.24

* For each language function the first row of numbers is the number of
messages in the specified group; the second row of numbers is the mean
for the specified group; and the third row of numbers is the standard
deviation of the specified group.

Table 3

Mann-Whitney Tests of Language Functions by Gender

Language     Asking       Giving       Presenting  Making
functions    information  explanation  opinion     suggestion

Mann         14443.50     14947.50     15040.50    15886.50
Whitney U
Asymp. Sig.      0.02 *       0.07         0.11        0.29
(2-tailed)

Language     Expressing
functions    disbelief

Mann         16387.50
Whitney U
Asymp. Sig.      0.94
(2-tailed)

Table 4

Mann-Whitney Test of Language Functions in First Messages by Gender

Language       Asking       Giving     Presenting    Making
Functions    Information  Explanation   Opinion    Suggestion

Mann          15385.50     14920.50     15402.00    16389.00
Whitney U
Asymp. Sig.       0.20         0.03 *       0.20        0.91
(2-tailed)

Language     Expressing
Functions    Disbelief

Mann          16183.50
Whitney U
Asymp. Sig.       0.56
(2-tailed)

Table 5

Descriptive Statistics of T-Units and Generation

Gender  T-unit   Generation

Male    248 *       249
          2.96        2.32
          2.26        1.33
Female  131         132
           3.65       2.03
           2.67       1.15

* For each group the first row of numbers is the number of messages in
the specified group; the second row of numbers is the mean for the
specified group; and the third row of numbers is the standard deviation
of the specified group

Table 6

T-Test of Messages in Terms of T-Units and Generation by Gender

              T    df   Asymp.Sig. (2-tailed)

T-units      2.71  377  <0.01 *
Generation  -2.05  379   0.04 *

References

American Association of University Women (1999). Gender gaps: Where schools still fail our children. Washington, DC: American Association of University Women Educational Foundation.

Aviv, R., & Golan G. (1998). Pedagogical communication patterns in collaborative telelearning. Journal of Educational Technology Systems, 26(3), 201-208.

Barrett, E., & Lally, V. (1999). Gender differences in an online learning environment. Journal of Computer Assisted Learning, 15, 48-60.

Bonk, C., Malinowski, S., Angeli, C., & East, J. (1998). Web-based case conferencing for preservice teacher education: Electronic discourse from the field. Journal of Educational Computing Research, 19(3), 269-306.

Crookes, G. (1990). The utterance, and other basic units for second language discourse analysis. Applied Linguistics, 11(2), 183-199.

Cookson, P. (1995, May). Analyzing interaction in audio-conferencing. A progress report. Paper presented at the Invitational Research Conference in Distance Education, Penn State.

Fennema, E., & Sherman, J. (1978). Mathematics achievement and related factors: A further study. Journal for Research in Mathematics Education, 9, 189-203.

Finocchiaro, M., & Brumfit, C. (1983). The functional-notional approach: From theory to practice. Oxford: Oxford University Press.

Graddol, D., & Swann, J. (1989). Gender voices. Oxford: Basil Blackwell.

Harasim, L. M. (1990). On-line education: An environmental collaboration and intellectual amplification. In L.M. Harasim (Ed.), On-line education: Perspectives on a new environment (pp. 39-64). New York: Praeger.

Harris, J., & Jones, G. (2000). A descriptive study of telementoring among students, subject matter experts, and teachers: Message flow and function patterns. Journal of Research on Computing in Education, 32(1), 36-52.

Henri, F. (1995). Distance learning and computer mediated communication: Interactive, quasi-interactive or monologue. In C. O'Malley (Ed.), Computer supported collaborative learning, (pp. 145-161), NATO ASI series, Berlin: Springer-Verlag.

Herring, S. (1992). Gender and participation in computer-mediated linguistic discourse. (ERIC Document Reproduction Service No. ED 345 552)

Herring, S. (1993). Gender and democracy in computer mediated communication. Electronic Journal of Communication, 3, 2.

Herring, S. (1994, June). Gender differences in computer-mediated communication: Bringing familiar baggage to the new frontier. Keynote talk at the American Library Association annual convention, Miami, FL.

Hunt, K.W. (1966). Recent measures in syntactic development. Elementary English, 43, 732-739.

Kaye, A. (1992). Learning together apart. In A. Kaye (Ed.) Collaborative learning through computer conferencing. Berlin: Springer-Verlag.

King, K., & McLeod, D. (1999). Coming of age in academic--A review of mathematics education as a research domain: A search for identity. Journal for Research in Mathematics Education, 30(2), 227-34.

Landow, G. (1992). Hypertext: The convergence of contemporary critical theory and technology. Baltimore, MD: Johns Hopkins University Press.

Li, Q. (1999). Teachers' beliefs and gender differences in mathematics: A review. Educational Research, 41(1), 63-76.

Lipponen, L. & Hakkarainen, K (1997, December). Developing culture of inquiry in computer supported collaborative learning. Paper presented at the second International Conference on Computer Support for Collaborative Learning. Toronto.

Machtmes, K., & Asher, W. (2000). A meta-analysis of the effectiveness of telecourses in distance education. The American Journal of Distance Education, 14(1), 27-46.

McDonell, W. (1992). Language and cognitive development through cooperative group work. In C. Kessler (Ed.), Cooperative language learning: A teacher's resource book. Englewood Cliffs: Prentice Hall Regents.

O'Donnell, A., & Dansereau, D. (1992). Scripted cooperation in student dyads: A method for analyzing and enhancing academic learning and performance. In R. Hertz-Lazarowitz & N. Miller (Eds.), Interaction in cooperative groups: The theoretical anatomy of group learning (pp. 120-144). Cambridge: Cambridge University Press.

Olsen, R. (1992). Cooperative learning and social studies. In C. Kessler (Ed.), Cooperative language learning: A teacher's resource book (pp. 85-115). Englewood Cliffs: Prentice Hall Regents.

Olsen, R., & Kagan, S. (1992). About cooperative learning. In C. Kessler (Ed.), Cooperative language learning: A teacher 's resource book (pp. 1-30). Englewood Cliffs, NJ: Prentice Hall Regents.

Scardamalia, M., & Bereiter, C. (1996). Student communities for the advancement of knowledge. Communication of ACM, 39(4), 36-37.

Scott, W., & Heller, R. (1991). Teamwork strategies for integrating women and minorities into the physical sciences. The Science Teacher, 58 (1), 24-28.

Shashaani, L. (1997). Gender differences in computer attitudes and use among college students. Journal of Educational Computing Research, 16(1), 37-51.

Stewart, C., Shields, S., & Sen, N. (1998). Diversity in on-line discussions: A study of cultural and gender differences in listservs. The Electronic Journal of Communication, 8(3) [Online]. Available: http://www.cios.org/www/ejc/ v8n398.htm

Tannen, D. (1994). Talking from 9 to 5. London: Virago Press.

Tannen, D. (1996). Gender and discourse. Oxford: Oxford University Press.

Wells, R. (1992). Computer-mediated-communication for distance education: An international review of design, teaching, and institutional issues. ACSDE research monograph, American Center for the Study of Distance Education. The Pennsylvania State University.

Yates, S. (1997). Gender, identity and CMC. Journal of Computer Assisted Learning, 13, 281-290.