ASSESSING WRITING BY ANALYZING WRITING:
ABILITY GROUPING AND IMPROVED STUDENT PERFORMANCE
Success in mastering news writing skills can be evaluated by students' handling
of both content issues (identification of the lead, selection and organization
of information, and accuracy in reporting facts) and mechanical issues (proper
use of English grammar and Associated Press style). This study examines the use
of ability grouping in a mass media writing course to improve writing students'
deficits in both content and mechanical skills.
Improvement in writing can be assessed by scores on tests of mechanical skills
or by evaluating a piece of writing itself. In this study, students' news
stories are content analyzed over a semester to measure changes in quality that
may be influenced by ability grouping. In addition, students' scores on grammar
and style tests are tracked throughout the semester to further assess the
effects of ability grouping on writing improvement.
Background
Journalism professors and professionals often have bemoaned journalism
students' lack of mastery in both content and mechanical skills. For example,
studies by Williams (1983), Stone (1990) and Auman (1995) reported professors'
and professionals' perceptions that journalism students' basic content and
mechanical skills were deficient.
To remedy these deficiencies, some journalism programs have required students
to pass grammar and language proficiency tests before enrolling in journalism
writing courses. John, Ruminski and Hanks (1991) reported that 36.4 percent of
the 236 journalism programs they surveyed required tests of English writing
skills. Adams (1978) reported that faculty in large journalism programs, which
enrolled 500 or more majors, generally believed that proficiency tests were
better used to pinpoint students' weaknesses than to exclude students from
courses. Some journalism programs have created writing labs or remedial
tutoring programs in which students work to improve deficiencies in mechanical
skills (Ryan and Pruitt, 1978; Hynes, 1978; Thayer, 1978; DiNicola, 1994), while
other programs have standardized the content of all basic writing courses as a
way to ensure students master a common set of skills (Blanchard, 1984). The
success of these techniques in improving student writing has been mixed (Hynes,
1978; Thayer, 1978; Blanchard, 1984).
In English writing courses at the elementary, secondary and college levels,
teachers long have experimented with proficiency testing and remedial measures,
including the use of ability groups, to improve students' skills. Ability
groups place students in groups with students of similar ability.
Studies that support ability grouping suggest students may perform better and
teachers may be more effective in addressing students who share similar
abilities (Oakes and Lipton, 1994). Some evidence, in fact, supports the use of
ability grouping to challenge higher-level students and to improve the
performance of students at all skill levels in high school (Newfield and
McElyea, 1983). On the other hand, ability grouping has been criticized for
potentially "ghettoizing" students by race, ethnicity or socioeconomic class,
which then impairs their self-esteem. According to Slavin and Braddock (1994,
p. 291), "Proponents of ability grouping have claimed that grouping is necessary
to individualize instruction for students and to accommodate their diverse
needs. ... In contrast, opponents of ability grouping have been concerned about
the negative effects of the practice on low achievers, in particular denying
them access to high- quality instruction."
Little research has been directed to the practice of ability grouping in higher
education, and particularly in journalism education. Haber's (1988) experiment
with within-class ability groups of students in a media writing course found no
statistically significant differences in the students' performance on a writing
skills test given at the conclusion of the course. However, Kanihan, Bunton and
Neuzil (1998) successfully used ability groups in a mass media writing course to
remedy deficiencies in students' mechanical skills. They found that placing
students into remedial and regular ability groups by scores on a placement test,
and then drilling students on mechanical skills, such as English grammar and
Associated Press style, produced an improved level of competence among remedial
students by the end of the semester.
The purpose of this study is to examine whether ability groups produce
improvements in students' writing content skills and to use ability grouping as
a strategy for improving students' mechanical skills. Borrowing from Suhor
(1985) and White (1989), this research looks at whether scores on the objective
exams used to measure mechanical competence would correlate positively with
scores on writing samples. Such a correlation would bolster evidence that using
ability groups is an effective way to teach journalistic writing.
Content skills in writing can be measured in various ways. In English writing
programs, for example, faculty have assessed writing samples with both
"holistic" and "primary trait" scoring (Odell, 1993). Holistic scoring insists
that writing samples be judged only on their overall meaning and presentation,
while primary trait scoring assumes that particular kinds of writing are
intended to meet specific goals and should be evaluated for their ability to
satisfy those goals (Odell, 1993, p. 297). In primary trait evaluation,
"readers judge written text in terms of characteristics necessary to accomplish
a specific rhetorical task" (Brossell, 1986, p. 180).
The study on which this paper is based used a set of scoring criteria for
evaluating news stories that borrows from the primary trait philosophy; the
study identified the traits that journalism faculty insist a well-written news
story should contain. For example, The Missouri Group's News Reporting and
Writing (1996) is a widely used reporting text that emphasizes such skills as
identification of the lead, selection and organization of information,
attribution of facts and quotations, and accuracy in reporting information.
Dodd, Mays and Tipton (1997) also insist on accuracy as a primary trait in news
writing; they used a survey to measure students' writing for improvement in
accuracy. Ryan (1995) presents a formula for news writing that emphasizes, as
does The Missouri Group, identification of key information for the lead, correct
attribution of information and use of quotations and facts to elaborate on
themes raised in the lead. All of these concepts were used to develop the list
of primary traits for evaluating students' news writing in this study.
This research also employs White's (1985) concept of the "anchor" or sample
paper, in which faculty use student writing samples to agree on the traits they
believe a well-written paper should contain. Anchor papers can be effective
means of making concrete the criteria for content analysis of writing, White
notes. "The goal is not only to obtain agreement on the scores of sample papers
and on the usefulness of the scoring guide but to help the readers internalize
the scoring scale by combining description with example" (p. 25).
Research Questions
This study asks two research questions: First, what is the relationship
between ability grouping of undergraduate mass media writing students and
improved competence in writing content skills, such as selection and
organization of facts, accurate use of information and identification of the
lead?
Second, what is the relationship between increased competence in mechanical
skills, such as grammar and Associated Press style, and improvement in writing
content skills?
Method
This study was conducted during the Fall 1997 and Spring 1998 semesters in
Media Writing and Information Gathering, an introductory course required for
journalism majors at a mid-sized, urban university in the Midwest. To examine
students' writing content skills, the researchers content analyzed news stories
written by each student at the beginning of the course, at midterm and at the
end of the course.[1] To examine students' mechanical skills, the researchers
measured performance on a grammar and style test given at three points in the
course. Subjects were students in the course; most subjects (N=73) were
sophomores and juniors (90 percent) who were majoring in journalism (63
percent).
Independent Variable
Ability group. Media Writing and Information Gathering students were placed in
ability group sections by scores on the Language Skills Test, an exam given
before course registration. The 75-point Language Skills Test covered grammar,
including agreement, redundancy, sentence structure, active verb use and
parallel structure. Students who scored 64 points or more were placed in
regular sections. Students who scored 60 to 63 points were placed in remedial
sections. Students who scored 59 points or lower were not permitted to enroll
in the course, but could re-take the Language Skills Test once; if they scored
60 or above the second time they were placed in a remedial section. A t-test
indicated significant differences (p<.001) between the remedial and regular
groups.
Procedure
In the Fall 1997 and Spring 1998 semesters, two remedial sections (N=25) and
four regular sections (N=48) of Media Writing were taught by five faculty
members. All subjects gave permission for a journalism department study of
writing courses but did not know the study's purpose. All faculty knew that
classroom data would be used for research purposes. Although this knowledge
might have created demand characteristics, any effects would have been similar
across all sections.
Dependent Variables
Mastery of content skills in media writing. Writing content skills were
measured three times during the semester in both the regular and remedial
sections. Students completed a baseline writing assessment on the first or
second day of the class. Faculty distributed a fact pattern about a fatal
traffic accident and asked students to write a newspaper story of up to 200
words in 30 minutes. At midterm, students were given the same fact pattern and
again asked to write a 200-word newspaper story in a 30-minute period. At the
time of the final exam, students were given an expanded fact pattern, which
included direct quotations and additional facts, and asked to write up to 250
words in 30 minutes. (The fact patterns are included in Appendix A.) Faculty
members did not assist students in writing any of these three news stories.
The stories were independently content analyzed by three researchers, who
identified "anchor" stories that exemplified writing competence and then
developed a set of criteria based on the anchor stories and the standard
concepts that journalism faculty emphasize in grading news stories in the
course. The criteria (included in Appendix B) emphasized skills in writing
content, such as selection and organization of facts, identification of the
lead, and accurate use of information. The criteria also included mechanical
issues, such as misspellings, sentence fragments and punctuation errors.
Each story, which totaled 75 points, was evaluated using these criteria.[2]
Competence level in mechanical skills in media writing. Students in both the
remedial and regular sections of Media Writing were drilled throughout the
15-week semester on grammar issues, such as parallel structure, subject-verb
agreement, dangling modifiers and active verbs, and Associated Press style,
including capitalization, abbreviations, numerals, punctuation and spelling.
Midterm and final exams, totaling 75 points each, measured competence level in
these mechanical issues using multiple-choice items. Each exam also asked
students to copyedit a news story for Associated Press style errors.
Results
Mastery of content skills in media writing
First, this research examined the relationship between ability grouping and
improvement in students' writing content skills, such as selection and
organization of facts, accurate use of information and identification of the
lead. These data reveal a positive relationship between ability grouping and
improvement in those skills. Students in the remedial ability group improved
their writing content skills to the point that their skill level was not
significantly different from students in the regular ability group by the end of
the course. (See Figure 1.)
At the beginning of the course, when all students wrote a news story of up to
200 words to serve as a baseline measure, the mean story score of the remedial
group was 39.12 and the mean story score for the regular group was 48.62
(p=.001), for a difference between means of 9.5. The standard deviation of the
remedial and regular groups at the time of the baseline story was relatively
high (11.45 for the remedial group and 8.00 for the regular group). This
indicates a fairly wide range of scores among both groups.
By the midterm point in the course, when the students again wrote a news story
of up to 200 words, the gap between the remedial and regular groups' scores had
begun to narrow. The mean story score of the remedial group was 52.96, as
compared to 59.50 for the regular group, for a difference between means of 6.54
(p<.001). This indicates the gap between students' scores had begun to shrink,
but a significant difference in their skills still existed. The standard
deviation of both groups also had decreased (5.47 for the remedial group and
6.80 for the regular group). This indicates that both groups of students
performed more consistently at the midterm than at the beginning of the course.
By the end of the course, when students were given additional facts and
quotations and asked to write a news story of up to 250 words, there was no
statistically significant difference in the mean story score between the
remedial and regular ability groups. The remedial group's mean story score was
48.82, and the regular group's mean story score was 52.28, for a mean difference
of 3.46. The drop in mean story score from the midterm to the end of the
course for both groups may be explained by the increased difficulty of the final
story assignment. However, despite the increased difficulty of the final story,
the remedial students were able to keep up with the regular students by the end
of the course. These results support a positive response to the first research
question of this study: Ability grouping leads to improved competence in
writing content skills. (See Table 1.)
For further analysis, the students' writing was broken down into two skill
sets: The first set of skills emphasized organization, including news judgment,
identification of the lead, selection of information and accurate use of facts;
the second set of skills emphasized overall writing skills, such as tone,
attribution, grammar, spelling and Associated Press style. (See Appendix B.)
At the beginning of the course, the difference in the remedial and regular
students' organizational skills approached significance (p=.06). The mean
number of organizational errors for the remedial group was 8.04, while the mean
number of organizational errors for the regular group was 5.87. The difference
in the two groups' organizational skills at midterm also approached significance
(p=.06), but the gap in mean number of errors was beginning to narrow. The
remedial group's mean number of errors was 4.96, while the regular group's mean
number of errors was 3.69. The standard deviations of both groups also
decreased from the baseline to the midterm. (See Table 2.)
By the end of the course, when more organizational errors were possible on the
longer final story than on the baseline or midterm stories, there was no
significant difference in the remedial and regular groups' organizational
skills. The remedial group's mean number of errors was 5.09, and the regular
group's mean number of errors was 3.64. The lack of significant difference in
the remedial and regular groups' organizational skills at the beginning of the
course stands to reason: None of the students could be expected to bring news
judgment skills to an entry-level media writing course. The continued lack of
significant difference -- at midterm and the final -- indicates that the
remedial students learned organizational skills at about the same rate as the
regular students.
At the beginning of the course, the remedial group's overall writing skills
were significantly different from the regular group's. The mean number of
writing errors for the remedial group was 27.92, while the mean number of errors
for the regular group was 20.51 (p=.001). By midterm, the gap in the number of
errors was narrowing. The mean number of writing errors for the remedial group
was 17.08 and 11.79 for the regular group (p<.001). The standard deviations of
both groups narrowed from the baseline to the midterm. (See Table 2.)
By the end of the course, no significant difference in the remedial and regular
groups' overall writing skills existed. The remedial group's mean number of
writing errors was 21.09, while the regular group's mean number of errors was
19.0. Although the mean number of writing errors for both groups increased from
the midterm to the final, the final assignment asked students to write a story
25 percent longer than before and included quotations and new facts, which gave
students more opportunity for error. However, the lack of significant
difference in the two groups' mean number of writing errors on the final story
indicates that students in the remedial groups were able to catch up with the
regular students.
Competence level in mechanical skills in media writing
This study also probed the relationship between ability groups and increased
competence in mechanical skills, such as grammar and Associated Press style.
Here, the data reveal a positive relationship. (See Figure 2.) On the 75-point
Language Skills Test used for placing students into remedial and regular
sections, the remedial group's mean score was 58.92, while the regular group's
mean score was 68.31 (p<.001). On the 75-point midterm mechanical skills exam,
the remedial group's mean score was 54.92, and the regular group's mean score
was 63.85 (p<.001).[3] On the final 75-point mechanical skills exam, the
remedial group's mean score was 66.18, while the regular group's mean score was
69.86 (p=.031). (See Table 3.) Although there are statistically significant
differences in the two groups' scores on the three tests, the mean differences
of all three narrow in the expected direction, from 9.39 on the Language Skills
Test to 8.93 on the midterm and 3.68 on the final. These results provide
additional support for the use of ability groups in media writing courses.
Relationships between mastery of content writing skills and competence in
mechanical skills
The second research question of this study examined the relationship between
improvement in writing content skills and increased competence in mechanical
skills. These data reveal a positive relationship between the skills. During
the course of the semester, instructors in the classroom were successful in
narrowing the gap between remedial and regular students' skills in writing
content, repeating the pattern found for improvement in mechanical skills.
These data support the construct validity of using objective exams to measure
writing ability. Correlations were computed for story scores and mechanical
skills test scores at corresponding time points throughout the semester; a
positive relationship between scores on the mechanical skills tests and the
ability to write a news story existed at all three points in the semester.
At the beginning of the course, the relationship between scores on the
mechanical skills test and the news story was significant (r=.38; p=.001). At
midterm, the correlation between the two scores also was significant (r=.58;
p<.001). At the final, the scores also correlated significantly (r=.39;
p=.026). (See Table 4.)
Discussion
The data in this study point to a positive relationship between ability
grouping and improved competence in writing content skills. This improvement
may have occurred because students in ability groups tend to share similar
skills or deficiencies, and instructors can therefore efficiently address those
issues in the ability group, as suggested by Slavin and Braddock (1994). In
mixed sections, instructors may have to spend more time addressing a variety of
student skills, while in ability groups, instructors may have the flexibility to
adjust the pace of instruction to students' needs, as indicated by Oakes and
Lipton (1994). That is, in a regular group, the instructor does not have to
spend instructional time on severe mechanical deficits, while in a remedial
group, the instructor may be able to spend more time coaching deficient students
on such issues as grammar and style. These results bolster the case for the use
of ability groups in journalism education made by Kanihan, Bunton and Neuzil
(1998).
The data in this study also suggest a positive relationship between increased
competence in mechanical skills and improvement in writing content skills, thus
providing a multi-method approach to support ability groups as a strategy for
improving student learning in mass media writing courses. In addition, as Suhor
(1985) and White (1989) suggest, correlating scores on objective skills tests
and writing samples provides a more sound means of evaluating student writing
than using one approach or the other.
The pattern of results presented in this study indicates the usefulness of
objective mechanical skills tests in placing students into ability group
sections, or perhaps even into the journalism major overall. After placement
into remedial groups by test scores on the objective mechanical skills test,
remedial students improve their writing content skills and mechanical skills to
the point where they can keep up with regular students.
The results also demonstrate correlation between writing content skills and
mechanical skills, suggesting that objective tests are valid as measurements of
mechanical skills and as positive indicators of students' ability to write a
news story. Thus, the results of this study establish the benefit for
journalism programs of using objective mechanical skills tests for entrance or
placement purposes, which would address the concerns of journalism faculty
reported by Adams (1978). Examining both writing samples and mechanical skills
test scores is the best-case scenario for entrance or placement purposes, but
these results indicate that journalism programs constrained by time or funding
may be able to use mechanical skills tests scores alone as a valid indicator of
student abilities. Mechanical skills test scores apparently point out "red
flags" in students' ability to succeed in the journalism major, or at least in
an entry-level writing course.
Future Research
This study addressed one of the directions for future research pointed out by
Kanihan, Bunton and Neuzil (1998), who suggested coupling a content analysis of
student writing with scores on objective mechanical skills tests to investigate
the relationship between these techniques and the use of ability groups. Future
research on the use of pedagogical techniques to improve student writing could
evaluate other types of writing -- advertising copywriting or broadcast scripts,
for example -- to see if placing students into ability groups in those areas
leads to improved writing skills. Student writing also could be tracked over
time. After students leave an entry-level ability-grouped course, how does
their writing fare? Can remedial students placed into mixed sections of
upper-level writing courses keep up with regular students?
Future research also could examine whether remedial students who are
"mainstreamed," or placed into mixed-ability classes, show the same kind of
improvement as remedial students placed into separate sections.
References
Adams, S.L. (1978). English proficiency among J-students. Journalism Educator
33, 1: 25-32, 53.
Auman, A. (1995). A lesson for instructors: Top 10 copy-editing skills.
Journalism Educator 50, 3: 12-22.
Blanchard, M.A. (1984). North Carolina standardizes new writing course
sections. Journalism Educator 39, 2: 18-22, 44.
Brossell, G. (1986). Current research and unanswered questions in writing
assessment. In K.L. Greenberg, H.S. Wiener and R.A. Donovan (eds.), Writing
Assessment: Issues and Strategies (pp. 168-180). New York: Longman Inc.
DiNicola, R. (1994). Teaching journalistic style with the AP Stylebook.
Journalism Educator 49, 2: 64-70.
Dodd, J.E., Mays, R.P., and Tipton, J.H. (1997). The use of an accuracy survey
to improve student writing. Journalism and Mass Communication Educator 52, 1:
45-51.
Haber, M.W. (1988). Do ability group courses bring results? Journalism
Educator 43, 1: 62-63.
Hynes, T. (1978). At California State (Fullerton), "EUT" proves its value.
Journalism Educator 33, 1: 34-36.
John, J.A., Ruminski, H. and Hanks, W. (1991). Trends in writing skills tests
for admission to programs. Journalism Educator 46, 3: 44-49.
Kanihan, S.F., Bunton, K. and Neuzil, M. (1998). Ability grouping in media
writing and the gap in mechanical skills. Journalism Educator, forthcoming.
The Missouri Group (1996). News Reporting and Writing (5th ed.) New York: St.
Martin's Press.
Newfield, J. and McElyea, V.B. (1983). Achievement and attitudinal differences
among students in regular, remedial and advanced classes. Journal of
Experimental Education 52, 1: 47-56.
Oakes, J. and Lipton, M. (1994). Tracking and ability grouping: A structural
barrier to access and achievement. In J. Goodlad and P. Keating (eds.), Access
to Knowledge: The Continuing Agenda for Our Nation's Schools (pp. 187-204).
New York: The College Entrance Examination Board.
Odell, L. (1993). Writing assessment and learning to write: A classroom
perspective. In L. Odell (ed.), Theory and Practice in the Teaching of Writing:
Rethinking the Discipline (pp. 289- 313). Carbondale and Edwardsville, IL:
Southern Illinois University Press.
Ryan, M. and Truitt, P. (1978). At West Virginia, the "DUT" points out
deficiencies. Journalism Educator 33, 1: 32-34.
Ryan, M. (1995). The use of a writing model for print or on-line news.
Journalism Educator 49, 4: 74-77.
Slavin, R.H. and Braddock, J.H. III (1994). Ability grouping: On the wrong
track. In J. Goodlad and P. Keating (eds.), Access to Knowledge: The
Continuing Agenda for Our Nation's Schools (pp. 289-296). New York: The
College Entrance Examination Board.
Stone, G. (1990). Measurement of excellence in newspaper writing courses.
Journalism Educator 44, 4: 4-19.
Suhor, C. (1985). Objective tests and writing samples: How do they affect
instruction in composition? Phi Delta Kappan 66, 9: 635-639.
Thayer, F. (1988). Remedial course for writing skills proven effective.
Journalism Educator 43, 2: 71-72.
White, E.M. (1985). Teaching and Assessing Writing: Recent Advances in
Understanding, Evaluating, and Improving Student Performance. San Francisco,
CA: Jossey-Bass, Inc.
________. (1989). Developing Successful College Writing Programs. San
Francisco, CA: Jossey-Bass, Inc.
Williams, H.L. (1983). Which are the toughest J-courses to teach ... and why?
Journalism Educator, 37, 4: 12-14.
Table 1
A Comparison of Writing Content Skills Competence
in Remedial and Regular Ability Groups
N Mean1 S.D. t Prob.2
------------------------------------------------------------------------
Baseline Story
Remedial Sections 25 39.12 11.48 -3.70 0.001
Regular Sections 47 48.62 8.00
Midterm Story
Remedial Sections 25 52.96 5.47 -4.45 0.000
Regular Sections 48 59.50 6.80
Final Story3
Remedial Sections 11 48.82 8.57 -1.13 0.274
Regular Sections 22 52.27 7.71
1 Each story totaled 75 points.
2 Two-tailed significance level.
3 N for the final story will increase at the May 1998 data collection point.
See note on p. 5.
Table 2
A Comparison of Organizational Skills and Overall Writing Skills
in Remedial and Regular Ability Groups
N Mean1 S.D. t Prob.2
------------------------------------------------------------------------
Baseline Story
Organizational Errors
Remedial Sections 25 8.04 4.81 1.97 0.056
Regular Sections 47 5.87 3.65
Baseline Story
Overall Writing Errors
Remedial Sections 25 27.92 8.65 3.71 0.001
Regular Sections 47 20.51 6.85
Midterm Story
Organizational Errors
Remedial Sections 25 4.96 2.64 1.97 0.055
Regular Sections 48 3.69 2.58
Midterm Story
Overall Writing Errors
Remedial Sections 25 17.08 4.54 4.38 0.000
Regular Sections 48 11.79 5.51
Final Story3
Organizational Errors
Remedial Sections 11 5.09 3.59 1.16 0.263
Regular Sections 22 3.64 2.99
Final Story3
Overall Writing Errors
Remedial Sections 11 29.09 6.79 0.84 0.412
Regular Sections 22 19.00 6.67
1 The mean represents the number of errors deducted from the 75-point story.
2 Two-tailed significance level.
3 N for the final story will increase at the May 1998 data collection point.
See note on p. 5.
Table 3
A Comparison of Mechanical Skills Competence
in Remedial and Regular Ability Groups
N Mean1 S.D. t Prob.2
------------------------------------------------------------------------
Language Skills Test
(Placement Exam)
Remedial Sections 25 59.92 4.25 -10.19 0.000
Regular Sections 48 68.31 2.47
Midterm Exam
Remedial Sections 25 54.92 5.99 -6.04 0.000
Regular Sections 48 63.85 6.02
Final Exam3
Remedial Sections 11 66.18 4.79 -2.45 0.031
Regular Sections 22 69.86 1.94
1 Each story totaled 75 points.
2 Two-tailed significance level.
3 N for the final exam will increase at the May 1998 data collection point. See
note on p. 5.
Table 4
Correlation Coefficients
for News Story Scores and Mechanical Skills Test Scores
Language Skills Test Midterm Exam Final Exam
Baseline Story .3749**
(N=72)
Midterm Story .5445*** .5765***
(N=73) (N=73)
Final Story1 .2762 .5593** .387*
(N=33) (N=33) (N=33)
Two-tailed significance level:
* = p<.05
** = p<.01
*** = p<.001
1 N for the final story will increase at the May 1998 data collection point.
See note on p. 5.
[1] NOTE TO REVIEWERS: At the time of the AEJMC paper deadline, a full set of
data for the fall 1997 semester (N=33 subjects) had been gathered, as well as
two of the three data collection points for the spring 1998 semester (N=40
subjects). There is no reason to expect that the pattern of results reported
here will significantly change after the third data collection point for the
spring 1998 semester occurs in May. A full set of data (N=73 subjects) and a
revised paper would be ready before the August AEJMC meeting.
[2] A check of intercoder reliability showed consistency among the three coders
(r=.87).
[3] The mean test scores for both groups dropped from the Language Skills Test
to the midterm because the midterm was a more difficult exam.
|
