4. Results and Discussion
4.1. Survey data on speech events
4.2.3. Statistical analysis of survey data on speech events
instructions, primarily utilise their English skills for oral and written translation (p.
531).
4.2.3. Statistical analysis of survey data on speech events
As described in Chapter 3, the bulk of both the tutor questionnaire and the student questionnaire, preceded by the three introductory questions seeking biographic data, was devoted to academic speech events, broken down into ten categories and organised around two dimensions: frequency and perceived importance. For the frequency dimension, respondents were asked to their evaluations on a four-member scale (‘always’, ‘often’, ‘sometimes’, ‘never’), subsequently converted into numeric values (4, 3, 2, 1) respectively. As Table 6 (Ranking of speech events according to the mean values for frequency measures based on the tutor responses (TF) and student responses (SF), respectively) demonstrates, the top-ranking speech event category from the point of view of frequency of occurrence as declared by tutors and students was associated with the statement The course description says that class participation will affect assessment, with a mean value of 2.89 and 2.28, respectively. This relatively high rating for the category ‘class participation’ may be accounted for by the comments attached to many of the electronically returned questionnaires by tutors, stressing the importance of oral involvement and contribution in the seminar courses. Another possible explanation in this regard may be sought in the relatively broad scope of oral activities the wording of the statement allows for, a circumstance likely to have affected the even more salient mean values for the perceived importance of this category in both samples (see Table 6).
108 It is also worth noting that the speech event category with the second highest mean in both samples was the oral presentation. It may be easily discerned from Figures 5 (Mean values for frequency measures based on tutor responses) and 6 (Mean values for frequency measures based on student responses) that, whereas in the tutor sample the prominence of the oral presentation is virtually eclipsed by three other categories, i.e.
small group work in class, note taking during class and students’ questions before/during/after class, the second place of the oral presentation seems to be unchallenged in the student sample.
Another noteworthy pattern seems to emerge from the comparison of frequency measures assigned by tutors and students: Although the distribution of debate, initiating/leading discussions in class and private conversation with the instructor is apparently varied on the basis of the responses supplied by the samples, both tutor and student respondents appear to be almost unanimous about the scarcity of team work outside class followed by an in-class report and interviews or professional consultations.
TF SF
class participation 2.89 class participation 2.28
presentation 2.28 presentation 2.11
small group work in class 2.20 note taking during class 1.73
note taking during class 2.13 small group work in class 1.72
students' questions before/during/after
class 2.12 debate 1.67
debate 1.86
students' questions before/during/after
class 1.61
initiating/leading discussion in class 1.74 private conversation with instructor 1.60 private conversation with instructor 1.44 initiating/leading discussion in class 1.59 team work outside class followed by a
report in class 1.38
team work outside class followed by a
report in class 1.07
interview/professional consultation 1.08 interview/professional consultation 1.00 Table 6: Ranking of speech events according to the mean values for frequency measures based on the tutor responses (TF) and student responses (SF), respectively.
109 Figure 5: Mean values for frequency measures based on tutor responses.
Figure 6: Mean values for frequency measures based on student responses.
110 By juxtaposing the ranking displayed in Table 6 with the one presented in Table 7 (Ranking of speech events according to the mean values for the importance dimension based on the tutor responses (TI) and student responses (SI), respectively), it becomes obvious that class participation is not only the most common feature but is also felt to be the most important oral element of a seminar course by teachers and students alike.
A major surprise, made visually palpable in the bar diagrams of Figure 8 (Mean values for the importance dimension based on student responses), is presented by the considerable advance of the category ‘debate’ in the student sample, rendering it equally important with class participation, whereas its relatively mediocre position is retained in the tutor sample.
TI SI
class participation 4.21 class participation 3.78
note taking during class 3.82 debate 3.78
students' questions
before/during/after class 3.64
initiating/leading discussion
in class 3.61
initiating/leading discussion in
class 3.61 presentation 3.56
presentation 3.58 note taking during class 3.50
debate 3.30
students' questions
before/during/after class 2.89 small group work in class 3.25 small group work in class 2.83 private conversation with
instructor 2.91
private conversation with
instructor 2.56
team work outside class
followed by a report in class 2.56
team work outside class
followed by a report in class 2.39 interview/professional
consultation 2.45
interview/professional
consultation 2.28
Table 7: Ranking of speech events according to the mean values for the importance dimension based on the tutor responses (TI) and student responses (SI), respectively.
111 Figure 7: Mean values for the importance dimension based on tutor responses.
Figure 8: Mean values for the importance dimension based on student responses.
112 As has been pointed out previously, for measuring the two dimensions of the survey two different Likert-scales were employed. Hence, whereas in the section inquiring about frequencies responses could be marked on a four-point scale, in the section meant to collect data on the importance dimension five options were offered: ‘extremely important’, ‘very important’, ‘fairly important’, ‘not really important’ and ‘completely unimportant’. To enable further comparisons between data collected about the same items but evaluated according to two semantically and numerically distinct scales, following the conversion of questionnaire responses into ordinal values (i.e. ‘always’ = 4, ‘often’ = 3, ‘sometimes’ = 2, never = 1; ‘extremely important’ = 5, ‘very important’ = 4, ‘fairly important’ =3, ‘not really important’ = 2, ‘completely unimportant’ = 1), partial two-tailed correlations were employed to establish correlation relations between the two dimensions within in each sample. In the tutor sample a strong positive correlation was discovered between frequency and importance values, with the correlation coefficient (r) determined at 0.682, significant at a p < 0.01 level of significance (df = 267). Values derived from the two data sets in the student sample, on the other hand, yielded a significant but relatively milder correlation value, the correlation value (r) being 0.365, significant at a p < 0.01 level of significance (df = 465).
In order to find out about correlation relations between the two data sets for individual speech event categories in the tutor sample and the student sample respectively, the Two-tailed Bivariate Spearman Test was carried for each speech event category. Table 8 (Correlation values for individual speech event categories based on the tutor sample and student sample, respectively) demonstrates that the correlation between frequency and importance values in the two samples was positive for all but
113 one category, namely ‘interview or professional consultation with an expert outside class’ in the student sample. Statistically significant correlations were detected in case of ‘small group work in class’, ‘initiating and leading discussions in class’ and ‘note taking during class’ at the 0.01 level in both samples, as well as ‘private consultation with the instructor’ at the 0.01 level in the tutor sample and at the 0.05 level in the student sample. For the categories ‘Presentation’ and ‘Debate’ statistically significant values were produced only the tutor sample.
speech event categories r/p correlation values for the tutor sample
correlation values for the student sample
class participation r
p
.218 .265
.054 .699 small group work in class r
p
.876**
.000
.379**
.005 team work outside class
followed by a report in class r p
.245 .284
.283 .070
presentation r
p
.871**
.000
.120 .389 initiating/leading discussion in
class
r p
.680**
.000
.508**
.000
debate r
p
.666**
.000
.139 .317 interview/professional
consultation
r p
.124 .687
- - note taking during class r
p
.552**
.002
.777**
.000 students' questions
before/during/after class
r p
.543**
.001
.087 .531 private conversation with
instructor
r p
.493**
.004
.358*
.016
* p<0.05
** p<0.01
r = correlation value; p = ‘p’ value
Table 8: Correlation values for individual speech event categories based on the tutor sample and student sample, respectively.
114 In addition, it is also worth noting that the second highest correlation value was found in relation to the OAP in the tutor sample, with a correlation coefficient of 0.876, significant at the 0.01 level. This discovery might suggest that instructors’ declared practice in terms of the OAP and the perceived importance of this genre seem to be in harmony. At the same, it also patently obvious that there is a considerable discrepancy between practice regarding the OAP as reported by participating students and the degree of importance they attributed to the OAP, with a rather low correlation coefficient of 0.120. With respect to interviews and professional consultations, a fundamental assumption was validated because the pertinent data showed zero variance.