CERG Technical Report 99/001 DRAFT

This CERG Technical Report covers the data collected on the 1998 first semester intake into SFT1101.

The first survey was carried out at week 8 and a similar survey was given at week 13.

The surveys were administered in tutorial groups and the Tutors entered the results for their groups.

This report covers the basic analysis modes on sample characteristics plus some regression and correlation analyses. Further analyses will be carried out to investigate structural characteristics and will be reported if the results are useable.

Total number sampled n=284

These categories were created through a content analysis of the open-ended responses.

The high rates for "Interest in computing" and "Career/job opportunities" are what would be expected from an undergraduate course.

These categories were created through a content analysis of the open-ended responses.

No relationship between hours worked and coping with course(Q.16).

These were mainly sport- based activities.

Note : 20 No Responses

Intercorrelations between the Class and Resource attitudes show no consistent pattern of relationships except for the obvious ones - lab classes by lab exercises, lectures by lecture notes etc.

From a regression analysis, the most significant impact on coping with the course from the attitudes to the course is the usefulness of the tutor. Lectures and assignments have the next highest impacts but far less than that for the tutor.

Nothing useful on attitudes

Nothing useful on study pattern

No difference on being first choice of course

Nothing useful

There were few differences based on gender.

Females are more likely to have had some programming study or experience before starting the subject. But all those (5 only) who indicated more extensive experience were males.

Females were more likely to see SFT1101 as more difficult than their other subjects and were more likely to say that they were not keeping up with their work in the subject.

Females tended to see their tutor as less useful than were males but the difference is marginal (0.2 scale points).

There were no differences on this breakdown of the data.

There was a marginal difference in how discussion classes were percieved with those working part time seeing them as less useful.

An uninterpretable result was found on the use of the SFT1101 Web site. It suggested that those doing part time work are more likely to be higher on heavy (5+) and low use (1-2) while those with no job will have higher medium (3-5) and non use (never).

Reported part time work hours have increased by the end of semester.

There is a marginal drop in hours worked from the first survey. This has to be balanced against the fact that some students in the first survey reported very large weekly hours.

The intercorrelations between the confidence variables show that students are relatively consistent in the way they rate the 3 items.

There is no significant change between mid-semester and end of semester on any of the attitude variables.

There is little direct relationship between confidence and the attitudinal variables.

The odd, interesting result is that there is an inverse relationship between confidence and Other students. That is, if you are confident you see other students as less useful and vice versa.

Tutors have no relationship to reported confidence.

A regression model was generated using the sum of the three confidence measures as the dependant variable. This model indicated that there was little potential causal relationship between the attitude variables and confidence.

Working on the assumption that the expected grading for the final result for SFT1101 could be seen as a quasi-continuous variable, the attitude variables were regressed on this. This regression model failed to produce an effective picture although there is some indication that the usefulness of assignments and lecture notes contribute to perceived course performance.

There is no relationship between tutor groups and the three confidence variables.

With the attitudinal variables, a similar pattern of results was found as for the first survey but this was confounded by the fact that two very small groups were consistently different from most of the rest.

In this survey differences did appear between fee paying status groups.

All confidence questions were rater lower by full fee students.

Full fee students saw the Discussion Classes as more useful but the result is practically marginal (0.24 of a scale point).

The full fee group used the SFT1101 Web page less than the other group.

There was less difference in this survey but the odd Web page useage result was maintained.

A difference which appeared in this survey was that Repeat Students rated the usefulness of Lecture Notes lower than first time students.