FROM STUDENTS' WRITING IN SECONDARY MATHEMATICS L Diane Miller, Curtin University of Technology One field of knowledge important to effective teaching is pedagogical content knowledge. With experience, teachers become aware of how students comprehend or typically misconstrue mathematical concepts, skills, and generalizations. They become aware of common misconceptions and "buggy algorithms" constructed by students and of the stages of understanding that students are likely to pass through in moving from a state ofhaving little understanding of the topic to mastery of it. It is important for mathematics educators to find ways to expedite the development of pedagogical content knowledge. The study reported in this paper suggests that in-class, impromptu writing prompts are one means by which teachers can get a clearer picture of students' understanding or nonunderstanding of mathematics in a relatively short period of time. For the past decade, mathematics educators and writing specialists have been combining their efforts and collaborating in research to examine the cognitive and affective benefits that students may directly derive as a result of writing in mathematics classes. Advocates have proposed various types of writing as being beneficial to students' development of higher order thinking skills, improving attitudes, and reducing anxieties. Journal writing (Nahrgang & Peterson, 1986; Borasi & Rose, 1989) has been shown to affectively benefit students by providing them an opportunity to express their anxieties toward mathematics and the problems they encounter in the learning process. Waywood (1989). concluded that journal writing can also cognitively benefit students by leading them to summarise and reflect on the mathematics they are learning. Expository writing (Bell & Bell, 1985) has proven to be an effective and practical tool for teaching problem solving. In general, Reuille- Irons and Irons (1989) suggest that writing activities "encourage children to be creators of their mathematical knowledge. When they create their own knowledge, they gradually build a picture of concepts and ideas that will be useful in problem-solving situations" (p. 98). (For an overview of the progress being made in a number of developmental projects on writing in mathematics classes, see Davison and Pearce, 1990.) Constructing Pedagogical Content Knowledge 31 But what about benefits that the teacher may directly derive as a result of reading students' writings? Data collected by Borasi and Rose (1989) suggest that teachers, too, may be equally affected in their teaching by reading their students' writings. There is a dearth of literature addressing the affect the use of writing in mathematics classes has on teachers; specifically on what teachers learn from reading their students' writings. The purpose of this paper is to partially fill this void by examining potential teacher benefits of using writing in secondary mathematics classes. Background to Current Research Several sources in the literature suggest that the expertise involved in cognitive aspect of teaching can be seen as the merging of at least three of knowledge. One of these is called lesson structure knowledge & Smith, 1985). Lesson structure knowledge includes the skills to plan and run a lesson smoothly, to create a smooth transition from Olue.les:son to the next, and to explain the material clearly. A second field is <" .. 1"".0,"1" matter knowledge (Leinhardt & Smith, 1985). No one can .if"" .,. need for a teacher to know the facts in a particular domain or the heuristics, and principles which operate on those facts. third field of knowledge important to effective teaching is called content knowledge. According to Shulman (1986), teachers possess "the understanding of how particular topics, principles, tegJleS, and the like in specific subject areas are comprehended or typically are learned and are likely to be forgotten" (p. 26). In like any other subject area, effective teachers should be aware of COflcelPtu.al and procedural knowledge that students bring to the learning of They should be aware of common misconceptions constructed by and of the stages of understanding that students are likely to pass in moving from a state of having little understanding of the topic to of it. teachers know that pedagogical content knowledge is, in C,QJnstructed over time through teaching experiences. For example, a fa,jjtlY;COJnrrlOn misconception developed early in the study of algebra is that 7 plus x rather than 7 times x. Experienced teachers know this and review the counterexample in their instructional comments to the correct meaning of 7x. The importance of pedagogical content to effective teaching is just beginning to arouse the interest of education researchers. The findings of Peterson, Fennema, and Loef (1989) suggest that teachers' pedagogical content may be importantly linked to teachers' classroom actions and, ultimately, to students' classroom learning in mathematics. The pdmary focus of pre-service teacher education programs is on subject matter knowledge and lesson structure knowledge. The beginning teacher's first concerns in the classroom are generally behavioural 32 Miller management and lesson preparation. Only with time does the construction of pedagogical content knowledge begin. If we accept the premise that pedagogical content knowledge is an important ingredient to effective teaching, then practicing teachers, mathematics educators, and researchers must ascertain how the construction of pedagogical content knowledge can be beneficially expedited. The results of a study conducted in the United States and reported in this paper, suggest that the use of in-class, impromptu writing prompts can promote the construction of pedagogical content knowledge for both experienced and inexperienced teachers. The Study In-class, impromptu writing prompts - the type of writing activity utilised in the study - are simply-worded statements or questions directing students' thoughts to the explanation of a single concept, skill or generalization. For example, a teacher may give the following prompt to a class of year nine students. "Why do we say that division by zero is undefined?" Prompts can also encourage students to express their attitudes and/or anxieties about mathematics or problems they encounter while learning mathematics. For example, "Tell me (in writing) one thing you like about learning mathematics and one thing you do not like about learning mathematics." Three secondary mathematics teachers from a large metropolitan high school in the United States agreed to participate in a study to examine the potential teacher benefits of using impromptu writing prompts in mathematics classes. This paper will focus on two of these teachers. One teacher was relatively inexperienced as a second year teacher; the other had more than fifteen years experience teaching secondary mathematics. They both taught first year algebra. Successful completion of first year algebra is a graduation requirement of the school district participating in this study. Generally, year 9 students (fifteen years old) enrol in first year algebra. One class of students (25-30 students per class) for each teacher participated in the study. Students were generally asked to write at the beginning of a class unless the prompt asked about something that occurred in class that day. They were given five minutes to read the prompt, formulate their response and write it on paper provided by the teacher. Initially students were asked to write during three out of every four instructional periods. As the study progressed, the number of writing tasks were reduced to about three per week. Students were not rewarded for writing nor penalized for not writing. Thus, every student did not write every day. Approximately fifty-five prompts were used in the study. Twenty were directed toward the affective domain; for example, attitudes toward and anxieties about mathematics, and feelings about class and topics being studied. Ten queried students' attitudes toward their responsibility in learning; that is, what determined whether or not they did their homework and how they prepared for an assessment. Twenty asked students to explain a Constructing Pedagogical Content Knowledge 33 mathematical concept, skill or generalization. Occasionally students were asked to write on a "fun topic" or to write anything they wanted to write on a topic of their choice. One prompt in this category asked students to identify their favourite single digit number and to explain why they liked it. Another prompt said "Write about any topic you want to today. Whatever you want to say, write it." Teachers generally read each set of writings the day they were collected. Once per week the teachers were asked to reflect upon what was being learned from reading the students' writings and to write about their predominant impressions. As the study progressed, the teachers' reflections became more evident in oral discussions than in their writing. The students' and teachers' writings were collected fortnightly for one sernes:ter along with notes taken during the collaborative team meetings and QllImg discussions with individual teachers. Since the focus of this discussion writing prompts can expedite the development of teachers' content knowledge, the remainder of this paper will concentrate prompts which asked students to write about their understanding content. . interpretive research methodology (Erickson, 1986) was Lerrlented in this study. Documentary materials consisted of students' to timed, in-class impromptu writing prompts, teachers' writings they were learning from the students' writings, and field notes from discussions with individual teachers and during meetings of the team. As suggested by Glaser and Strauss (1967), a thorough /f!j:][ftityrSis of the students' writings was completed. These writings were for recurring patterns that contributed to the identification and <i. .... O:< .. ..."' of ways in which impromptu writing prompts enhanced the ability to assess how well students were comprehending the ",/U,;;O;'DU they were studying. In many respects, this was an exploratory .;; </?,';.....L$fi1(1y very little research has been undertaken to examine benefits for >ii;;K;';.)Uj in using writing in mathematics classes. Thus, it was deemed ..... ;..... "" ...... /jJiapprclprilate to search for conclusive empirical evidence which would allow >tt'lleilrs,eat'chers to predict teacher benefits from using impromptu writing A;!;'i.;"t;'/.; L"r>tlomlT>ts in a mathematics class. More relevant to the research team was an of what teachers can learn from reading their students responses tQIIllpf()m'ptu writing prompts. What Three Prompts Yielded The Commutativity Prompt Students sometimes view mathematics as isolated parcels of information which must be memorised for "the test" and then forgotten so that new parcels can be memorised and the cycle continue. Mathematics is a study ofrelationships, but is often presented as independent packets of knowledge which are not related to other mathematical concepts or "real world" 34 Miller applications. Sometimes teachers fail to provide experiences which help students construct mathematical relationships, other times their attempts are not successful. Evidence of a failed attempt surfaced when the first year algebra students in this study were asked to write about their understanding of commutativity. When implementing an interpretive research methodology in an educational setting, the contextual background is crucially significant to the sense-making process of the data being examined and conclusions drawn. The following paragraph will provide the reader with the background for this particular prompt. Three months into the school year, an assessment asked students to demonstrate an understanding of the Commutative Property for Addition and the Commutative Property for Multiplication by stating the properties in abstract form and labeling statements that represented each property; for example, students were asked to identify the property represented by 5 +7 = 7 + 5. A majority of the students correctly quoted the rules in their abstract form (for example, for real numbers a and b, a +b ::: b + a and a x b = b x a) and labeled statements as the Commutative Property for Addition and the Commutative Property for Multiplication. During a team meeting, someone asked if the students would be able to apply what they knew about commutativity to operations other than addition and multiplication. The two first year algebra teachers indicated that they had briefly commented on division and subtraction in class and that exercises in the students' textbook had also provided counterexamples using division and subtraction. The team decided to ask the students to respond to the following prompt. You have studied the Commutative Property for Addition and Multiplication for real numbers. Not all operations are commutative. Explain why each of the following operations are not commutative: (1) Division (2) Subtraction (3) Raising a number to a power The teachers were confident that not all students would produce an acceptable response. However, they were equally confident that the majority of their students would respond acceptably. Fifty papers were collected from two classes. The students' responses are listed in Table 1. Four of the fifty students (8%) produced acceptable responses to all three operations. (Marked acceptable by the teachers, two research assistants, and a mathematics education researcher.) Eleven (22%) acceptably responded to division and nine (18%) demonstrated an understanding of why subtraction is not commutative. The teachers were most disappointed in the students' responses to division and subtraction. While the students had worked with numbers raised to a power and had adequately demonstrated they knew what aX meant, the teachers had not discussed the operation in terms of being noncommutative. The query raised in the prompt asked students to relate prior knowledge; that is, their understanding of addition and multiplication being commutative, to an unfamiliar context, raising a number to a power. Constructing Pedagogical Content Knowledge 35 Table 1 Responses From Students Responses deemed acceptable by the research team 1. Division - (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) (k) division is not cummutative because 3 + 5 is not equal to 5 + 3 because you can't switch them around because you will come out with a wrong answer because they don't have an order and you can't change them around Because you can't change them around _ you cant change the numbers around 3)450 :t: 450)3 see in addition you can change the order 3 + 5 = 5 + 3 multi 3 x 5 = 5 x 3 It just wont work out.2... --.5. 3)6 but 6)3.0 Because in division you come up with a different answer you can't reverse the numbers and come out with the same answer isn't cummutative because you can't switch numbers around to make it easier 8 4 -:t:- 4 8 Because if you have 10 + 5 = 2 you cant say 5 + 10 because t ~ t wQ!k out if you change the numbers around the answer will change 8)10 10)8 2. Subtraction - (a) subtraction is not commutative because 3 - 2 is not equal to 2 - 3 (b) same for division if you swich them around and try to subtract your going to come out with a wrong answer (c) has to be in certain order (d) They have to be in a certain order (e) It doesn't work out either 6 - 3 = 3 but 3 - 6 = -3 (f) the number can only be done one way to come out with that answer (g) when you subtract if swich the numbers you get a different number 8 - 5 = 3 5-8=-3 (h) Because if you tum a number around like 3 - 2 = 5 it won't work out as 2 - 3 (i) If you change the numbers the answer will change 5 10 - 10 -....5. 3. Raising a number to a power- (a) raising a number to a power is not cummutative because 3 5 is not equal to 53 (b) cant example 3 2 = 9 2 3 = 8 (c) Because if you have a number like 2 3 = 8 it will be different if you had 3 2 = 9 (d) If you change the number the answer will change 53 3 5 Responses deemed unacceptable by the research team: 1. Division - (a) a + b + c = a a + b + C just change the variable around (b) A+B =B+A (c) 1 7)7 1 o Cd) commutative is not a division because the rule don't say it (e) ~ 6)6 Q o (f) because you divide (g) it doesn't work L -...5. 4)8 4)2.0 36 Miller Table 1 continued 3. Raising a number to a power - (a) rrrrrrr =r 7 (b) the same as above (see division d. & subtraction d.) (c) 7 7 aaaaaaa (d) Because you don't (e) doesn't work. I don't know how if it works (t) +4 3 (g) It just doesn't work. I don't know why. It just wont. (h) 2 3 (i) God made it that way G) you use power (k) 5 2 '2=2'5 2 I don't understand why the following operations are D.Qt commutative 48 + 6 72+8 3 + 5 =5 + 3 5 3 .i.= 5'5 3 102'100 =100102 you can't divide 4)2 and get a whole number .2 4)8 GQd made.iUhat way 4)8 = 4)8 Because commutative property has nothing to do with division
5)3!L6)30 _ 10)100 1(0)10 For division if you divide the larger number into the smaller number the answer will not be even (ex. -2Q 10)900 2Q o Q but ifit is done the other way, you have to add dismials (ex. .0111 9(0) 10.0000 1000 200 1000 -900 1 (h) (i) G) (k) (1) (m) (n) (0) (P) 2. Subtraction - (a) a - b - c = a + b - c just change the variable around but subtract (b) A-B =B-A (c) 3 - 4 =1 (d) commutative is not a subtraction because the rule don't say it (e) 2 - 1 =1 (t) Because you subtract (g) 5 - 3 = 2 doesn't do it either maybe 5 - 3 = -2 (h) ex. 12 - 6 =6 - 12 =6 It wouldn't work out like that because you can't subtract 12 from 6 (i) 4-3=3-4 G) 4 -2 2 (k) God made it that way (1) 6 - 12 = 12 - 6 because you can't subtract 12 from 6 (m) Because you subtract differently than adding (n) you subtract (0) 7 - 2 = 2-7 (P) 7 - 2 =5 5 = 2 - 7 Constructing Pedagogical Content Knowledge 37 The students' writings to this prompt contributed to the development of these teachers' pedagogical content knowledge. While the majority of these students had successfully quoted rules in abstract and had recognized their use in specific examples, they did not really understand the concept of commutativity. The experienced teacher expressed her disappointment but was not shocked by the results. The second year teacher was both disappointed and surprised by the results. In one team meeting, the second year teacher kept reiterating that she had discussed this in class and their textbook had given them counterexamples using division and subtraction. generally, a teacher cannot assume that because students can state rules in the abstract that they really understand the principle or can apply it to an l1t1familiar context. This may seem like a trivial, universally known fact in but how often do teachers proceed with the curriculum assuming 1:P(lents know more than what they really do. Seventy-eight percent of the fifty students returned unacceptable responses, blank papers, or papers with know" written beside each query. Clearly a majority of these students not relate what they knew about commutativity of addition and rI11.l1tiplication to division, subtraction, and raising a number to a power. . Other than being reminded that students can quote rules without C9111prehending their content, the teachers' pedagogical content knowledge al$qoenefited in other ways. For example, look at the students who made about basic facts in subtraction: 5-3 = -2; 12-6 = 6-12 = 6; 3-4; 7-2 = 2-7. Once again, the inexperienced teacher was surprised by errors. Her comments indicated that she believed her students knew that $-3was not -2; that 6-12 was not 6; that 4-3 was not equal to 3-4; and that 7-2 equal to 2-7. But, she had no explanation why these statements in their writing. A comment made to the mathematics education was "It's like they're not thinking about what they're writing. just putting down things that don't make sense." Perhaps this is an of students manipulating symbols and not thinking about what these mean in the process of trying to make sense out of something new to them. Very little is known about the processes of assimilation and .. aommodation as students are constructing new knowledge in mathematics. .Many of the students' statements were nonsense; but was that because their !houghts were in disarray as they were trying to relate prior knowledge to a new context and make it meaningful? In the short term, these teachers decided to return to commutativity, emphasizing its meaning and reillustrating wl1Yit was applicable to some operations and not to others. In the long term, they each commented that future lessons on this topic would be revised to emphasize the meaning of commutativity rather than the memorization of an abstract rule. The Ratio Prompt A second prompt that was beneficial to the construction of the teachers' pedagogical content knowledge was the following: . 38 Miller In working the following problem, write down not only everything you do, but everything you think about in coming up with your answer. A painter uses four cans of paint to paint six rooms. How many rooms can be painted with six cans of paint? Once again, the contextual background for this prompt is helpful in interpreting what the teachers learned from reading the students' writings. The first year algebra classes covered ratio and proportion early in the academic year (September). Several problems had been worked where students had solved for an unknown term in the proportion. A few word problems had been solved in class and assigned for homework, but the general assessment had provided students with a structured proportion with a missing term and the students were asked to solve for the unknown; for example, :z =. The preceding prompt was given in late October. The teachers were wanting to ascertain how many students would use a proportion to solve the problem. Forty-eight papers were collected. Ten (21 %) were returned either blank or with "Don't know" written beneath the prompt. Twenty students (42%) submitted unacceptable responses. Fifteen of these twenty said the correct answer was eight. Six students submitted eight as an answer but provided no explanation for their solution. Nine of these fifteen students provided a rationale similar to the following response: Because 6 is two more than 4, I will add 2 to 6 and get 8. 6 cans of paint will do 8 rooms. Eighteen students (37%) submitted an acceptable response. Six of these 18 said the answer was 9 but provided no explanation for their solution. One student wrote the following: 4 cans for 6 rooms 6 cans for 9 rooms I looked for something half way between. 8 cans for 12 rooms Seven students answered 9 as the number of rooms and provided an explanation similar to the following: Half of 4 is 2. 4 cans can paint 6 rooms. 2 cans can paint 3 rooms. If you have 6 cans, you are only adding 2 more to 4 so 3 +6 =9 Four students answered 9 with the following work submitted: Constructing Pedagogical Content Knowledge 39
4)6 No one used a proportion to solve this problem. Rather than using a solution process which was relatively new to the students, they had chosen to rely on othermeans of solution which were more meaningful to them. . ... The teachers' pedagogical content knowledge benefited in two ways from the students' writings to this prompt. First, they realized that students to adopt new methods of solution when their prior knowledge 3ll d experiences work for them; and second, the writings support the existence of a misconception or buggy algorithm which several students a,pplied (for details on 'buggy' algorithms see Davis, 1982; Davis, Jockusch, 1978). That is, when one number is increased by two, the number in the problem is increased by two. Once again, the teacher was not surprised by this error while the inexperienced was. Both teachers indicated that future lessons on solving problems gf:thistype would include comments about the buggy algorithm. The Variable Prompt The third prompt presented in this paper provides additional examples misconceptions in first year algebra. The teachers had introduced tlle90ncept of variable early in the year and had emphasized that any letter can lJ/llsed to represent an unknown quantity in a problem. They were satisfied \\lith students' ability to translate simple statements into mathematical for example: "Write a mathematical sentence for the following: Jft9Pf times a number equals twenty-eight." The majority of the students were writing 4x = 28 or 4n = 28, etc. The following prompt was given in mid-November, three months into the school year: Pretend you are trying to explain why "x" in the first and "y" in the second equation will be the same number. Pretend you are trying to explain this to a ten year old. Write how you would go about it. 4x=28 4y=28 Twenty-five papers were collected from the second year teacher's class. Four blank papers were returned. Four students submitted an acceptable response. The following response reflects the general statements made by these four students. "If you multiply 4 by x, you get 28. If you multiply 4 by y, you get the the same answer. Since you are multiplying by the same four, x and y have to be the same number because you get the same answer." Four other students attempted an answer, but submitted unacceptable responses. For 40 Miller example, one student wrote "Because you can use any variable in an equation as long as you keep everything else straight along in the equation." Two students submitted responses which were classified as tautologies by the researchers. One student wrote "x & y could stand for anything but they stand for the same thing here." Four students wrote what the researchers classified as nonsense; one example being, "I tell hem or her to read it." Seven students submitted responses that proved enlightening to the inexperienced teacher but did not surprise the experienced teacher. During a team meeting, the experienced teacher indicated that she knew from cumulative years of experience that students made these types of errors. These seven responses, as written by the students, are listed below. A discussion of their significance follows. (a) All letters of the alfebeat are = to 1 that is why they are the same number. (b) Any letter is I: S is 1, A is 1, B is one, Cis 1. So x and y will be one too. Any letter will be one. (c) The x and yare the same because they are alpabets and all alpabets equals the same. ~ x = 7 y = 24 Since the 4 and 28 are the same you divide to fmd x and subtract to find y because x comes before y. (e) 4x = 28 x is the ungiven number so you subtract 28-4 x=24 (f) If 4x = 28 x = 24 and 4y = 28 y = 24 Both of the problem is the same just y and x will = the same (g) I would say that you have to find out what x and y is. The first two students clearly have the misconception that any letter of the alphabet is equal to one. These students did not sit near one another so the teacher concluded that they arrived at this misunderstanding independent of each other; that is, they did not copy prompt responses in class. The next example (c) is interesting because it alone suggested that the student believed all letters in the alphabet were equal or represented the same number. One interpretation of this response is that x and yare the same in these two equations: 5x = 30 and 3y = 9. This turned out to be an incorrect interpretation. The teacher talked to this student during seatwork one day and learned that the student was trying to explain that when everything else is the same, then any letter in the alphabet can be used to represent the unknown number. At this stage, the student was much better able to communicate mathematical knowledge verbally than in writing. However the teacher's discussion with the student was initiated by the student's writing, thus giving credit for the growth in pedagogical content knowledge to the writing experience. The responses of the next three students suggest that they think the relationship between a coefficient and the variable is additive/subtractive. Example (d) reflects an attempt at justifying the interpretation, although not Constructing Pedagogical Content Knowledge 41 Discussion is provided to attempt an analysis of the student's thoughts. The student submitting (D is consistent with the misconception. The last example suggests that the student is not ready to accept that x and yare the same. The provided is to find out what they are, but no attempt is made to solve the equations. The inexperienced teacher found these misconceptions surprising. On her traditional types of assessments when students were asked to solve for x pry in an expression like 4x =28, she had assumed an answer of 24 had been .. a."careless error." Her own writings summarized her feelings: "The writings gave me an opportunity to look at each students' work individually. 1learned that the words "yes" and "I understand" are not precise answers. Students thought they understood but their writing indicated that they did .t10tunderstand." ..... .... . The experienced teacher was not surprised at the misconceptions which SMrfaced in the students' writings. The writings were beneficial to the teacher because they helped her to assess which individual held what misconceptions more quickly than using traditional whole discussions. An excerpt from the experienced teacher's writings amply this assertion: "For me as an experienced teacher, I found that I got picture much sooner of how the students were thinking I was able to locate specific problem areas and make corrections that I may not have been aware of had we just solved Fieldnotes from several team meetings indicated that the teacher was having some of her pedagogical content knowledge by the students' writings while the inexperienced teacher was using students' writings to help construct her pedagogical content knowledge. .............. Of the twenty prompts designed to assess individual students' of mathematical concepts, only one proved nonbeneficial to the of the teachers' pedagogical content knowledge. It asked StP:gents to copy a problem worked in class during seatwork or at home and "step by step in your own words" what was done to arrive at a final Inevitably, students copied problems "step by step" using only mbols with no explanation in words about the process they used or the thoughts they had while working the problem. The inexperienced teacher qyd this prompt six times until one student wrote: "To be fully/totally absolutely honest w/you. 1m sick of this [prompt]." The teacher stopped u.sing the prompt. Space does not permit a detailed analysis of all the prompts used in the study. The results of other prompts are detailled in Miller (1990), Miller (in .press) and Miller and England (1939). 42 Miller Much of what was learned by the teachers from reading the students' writings served to reconfirm pedagogical content knowledge previously constructed by the experienced teacher while supporting the development of new pedagogical content knowledge for the inexperienced teacher. For both teachers, the writings helped to identify individual students' misconceptions and nonunderstanding much more quickly than through the traditional whole class discussion process. Questions raised by critics of the study have been "What's so unique about impromptu writing prompts? Can't teachers learn the same things by asking students questions in class and by putting prompts of this type on assessments?" While teachers can ask open-ended, thought provoking questions in class or pose them in a context of small group activities, the general scenario is that only a few students become involved in these discussions. Allowing sufficient wait-time for students to formulate and respond to open-ended questions orally can create a lull during which other students can become off-task mentally if not physically, too. Within the time allotted for teacher-student interaction in class, a teacher cannot usually engage every student in a class of 25-30 students. A unique feature about impromptu writing prompts is that each student has the opportunity to express themselves, in writing, to the teacher during every class in which a prompt is used. Writing also allows them the opportunity to examine their thoughts and make changes in their statements prior to submitting their paper. Students answering orally generally do not have the chance to say something, reflect upon what they said and then make changes to those statements. The teachers in this study, found that five minutes was sufficient time for students' reading and responding to a prompt, but was not disruptive to the continuation of their other instructional practices. Because their classes had a range of academic abilities including students who were being mainstreamed from special education classes as well as those intending to go to college, the teachers did not want to reward students for writing or punish them for not writing. Some students had great difficulty expressing themselves in writing, particularly at the beginning of the study. While one might argue that students should be able to communicate their understanding of mathematics orally and in writing, the teachers in this study did not want to handicap or embarrass students by using writing prompts as official assessment tasks. It should also be noted that this strategy of feedback is not the panacea for assessment that some look for, but an effective additional strategy to other proven procedures. But it too has its limitations. As noted in the results section, on a few occasions some students gave inadequate or ambiguous written statements which needed verbal clarification by the teacher. However the subsequent verbal discourse, guided by the student's written response, zeroed onto particular difficulties much more quickly than would have been the case without this guidance. As previously stated, this study serves as a beginning to the examination of teacher benefits to the use of writing in mathematics classes. For the second year teacher in the study, the benefits were more pronounced Constructing Pedagogical Content Knowledge 43 than for the experienced teacher. Both continued using writing prompts during the second semester of the school year even though the study officially ended with the first semester. Teachers and mathematics education researchers should begin to investigate further the understudied phenomenon of benefits gained by the teachers who use writing activities in the mathematics class. The reconfinnation and construction of pedagogical content knowledge is only one potential benefit. Other benefits worthy of investigation are the influence reading students' writings can have on instructional practices, both short- and long-term; and, the influence reading students writings can have on teacher- student interaction patterns. The results of this study suggest a relatively simple, easily implemented, and fairly robust method for assessing students' understanding of mathematics at very little cost to either teachers or students. There is a need for a more controlled study measuring and documenting the teacher benefits suggested in this paper. The results could be very significant and influential to the professional development of novice and expert teachers. References Bell, E. S. & Bell, R. N. (1985). Writing and mathematical problem solving: Arguments in favor of synthesis. School Science and Mathematics, 85(3), 210-221. Borasi, R. & Rose, B. J. (1989). Journal writing and mathematics instruction. Educational Studies in Mathematics, 20,347-65. Davis, R. B. (1982). The diversity of errors in mathematics. Journal of Mathematical Behavior, 3(2), 73-77. Davis, R. B., Jockusch, E., & McKnight, C. (1978). Cognitive processes in learning algebra. Journal ofChildren's Mathematical Behavior, 2(1), 10- 230. Davison, D., & Pearce, D. (1990). Perspectives on writing activities in the mathematics classroom. Mathematics Education Research Journal, 2 (1), 15-22. Erickson, F. (1986). Qualitative methods in research on teaching. In M.e. Wittrock (Ed.), Handbook of research on teaching (pp. 119-161). New York: Macmillan. Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualitative research. Hawthorne, NY: Aldine Publishing Company. Leinhardt, G., & Smith, D.A. (1985). Expertise in mathematics instruction: Subject matter knowledge. Journal ofEducational Psychology, 77, 247- 271. Miller, L. D. (1990). When students write in algebra class. The Australian Mathematics Teacher, 46 (2), 4-7. Miller, L. D. (in press). Writing to learn mathematics. Mathematics Teacher. 44 Miller Miller, L. D., & England, D. A. (1989). Writing to learn algebra. School Science and Mathematics, 89 (4), 299-312. Nahrang, C. L., & Peterson, B. T. (1986). Using writing to learn mathematics. Mathematics Teacher, 79, 461-465. Peterson, P.L., Fennema, E., & Carpenter, T.P. (1989). Teachers' pedagogical content beliefs in mathematics. Cognition and Instruction, 6 (1), 1-40. Reuille-Irons, R., & Irons, C.J. (1989). Language experiences: A base for problem solving. In P.R. Trafton & A.P. Schulte (Eds.), New directions in elementary school mathematics (pp. 85-98). Reston: National Council of Teachers of Mathematics. Shulman, L.S. (1986). Paradigms and research programs in the study of teaching: A contemporary perspective. In M.e. Wittrock (Ed.), Handbook ofresearch on teaching (pp.3-36). New York: Macmillan. Waywood, A. (1989). Mathematics and language: Reflections on students using mathematics journals. In R. P. Hunting (Ed.), Language issues in learning and teaching mathematics (pp. 123-139). Bundoora: La Trobe University School of Education.