• No results found

Results.

Rationale for each stage of analysis.

The analysis of the data collected comprises two basic stages, validation of implementation of experimental lectures and testing of experimental hypothesis.

Validation of Study.

The first section of the analysis looks at the implementation of the experimental lectures. This process also can be broken down into two stages; the analysis of control and presentation style of lectures; and the analysis of OHPs and handouts. The analysis of control and presentation style. The analysis here is seeking to discover if the type of control (Lecturer, Student) and presentation style (Analytical, Holistic) were effectively manipulated during the experimental lectures, for example were analytical lectures rated as such by participants? The analysis also seeks to discover if the four lectures of each type were presented uniformly, for example was the lecturer controlled analytical lecture presented to the psychology group the same as that presented to nursing students? It is

important that the lecture5 formats were stable so that comparisons between the lectures can be made.

The hypotheses presented are:

1. The experimental lectures type of control and presentation style will be rated in line with the experimental design.

2. There will be no significant difference in the rating of the same experimental lectures presented to the different degree course groups.

The analysis of OHPs and Handouts: The analysis here is undertaken to further confirm the uniformity of the general features of the presented lectures. That is to

verify that what differs between the lectures is the type of control and presentation style and not the content of the OHPs or handout presentation style, for example. If these general features did in fact vary it would be difficult to make any

conclusions about possible differences in the responses to the experimental lectures; since the source of the variation would be unclear.

The presented hypothesis states:

1. There will be no significant difference in the rating of lecture OHPs and handouts within or between the experimental lecture types.

Testing of experimental hypothesis.

The analysis of effect of experimental lectures on perceived learning: Three questions from the data collection address this issue, effect of control, effect of style and the overall rating of the lecture. The first two questions are designed to give an insight into the effect of the specific features of the experimental lectures. The final question is designed to gain an insight into the general impact of the style of lecture on the participants’ perceived learning.

The hypothesis presented states:

1. There will be a significant difference in the rating of the effect of lecture control on perceived learning between the five learning styles.

2. There will be a significant difference in the rating of the effect of lecture style on perceived learning between the five learning styles.

3. The most positive overall rating of a lecture will be gained from the learning style which is attuned to that lecture.

17

Imputing— of data.

Part of the experimental design of this study was to exploit ‘real’ lectures which were part of the participants’ actual degree programme. This allowed the study to present lectures which had true saliency for the participants and therefore would be attended to by the students with attention comparable to any other degree course lecture. Such saliency would be lost with laboratory style lectures, creating doubt concerning the applicability and generalisability of data. Although bringing considerable rewards such an experimental design also has its problems. The main problem as far as this study is concerned is that the students were free to attend

17 Imputing is used to refer to the way in which missing data was delt with for analysis.

the lectures as they would any other lecture on the course (lecture attendance is not compulsory on the degree programmes). This meant that not all students attended all lectures; making the lectures compulsory would instantly destroy the ‘naturalistic’ design of the study. Analysis of the data reveals the following attendance across the four experimental lectures.

36 Participants attended all four experimental lectures. 41 Participants attend three experimental lectures. 33 Participants attended two experimental lectures.

8 Participants attended one experimental lecture.

If analysis of the data was to use repeated measures techniques, it should only be undertaken on those participants who attended all lecture conditions. Attendance at all lectures for this research was 36, it was felt that this figure was small and when divided down into degree courses became exceptionally small (Only four psychology students attended all lectures). For this reason it was decided that data would be imputed for those participants who attended three lectures; this would establish a population of seventy seven. This established a reasonable number of participants from each degree course; 35 Physiotherapy students, 14 Nursing students, 20 Statistics students and 8 Psychology students

(N.B. this procedure meant that 16.56% of the total data used within the analysis would be imputed).

Procedure for Imputing data.

The data could be imputed in two ways,

1. From the mean result of that question for the associated degree course, e.g. a missing result for rating of control of a Psychology student would be imputed from the mean control rating for the psychology group.

imputed from the mean of that students’ ratings of control for the other three experimental lectures.

(The third possibility of imputing data from Teaming style means’ was rejected as this would mean that the possible differences in lectures presented to different degree groups would be ignored, as all learning style were present in all degree groups. This could easily lead to type one error when analysing stability of lecture presentation styles or types of control).

After considering the available methods of imputing data it was decided that the method that would be used would be that which used the individuals’ scores for the rating of that category of questions. This conclusion was reached because this method would be the most stringent on the experimental hypothesis, reducing the

18

chance of a false positive (type one error) being established .

18 Analysis of the data using the degree course method of imputing data was undertaken. The results revealed no relevant differences to those found when the individual method of imputing data was used. This analysis is not reported here.