• Keine Ergebnisse gefunden

2 Future Time Perspective (FTP)

9.1 Investigation 1

9.1.1 Effects of Instruction 169

9.1.1.1 Motivational Beliefs

There was only one dependent variable showing significant differences at the group level of comparison: control beliefs (p < 0.05). Closer examination of the pairwise comparisons revealed that students receiving future oriented (full) instruction reported higher levels of control beliefs than students in the non-future oriented (modified) and non-instruction (control) groups, however the differences in variation were not significant.

To follow-up on the MANOVA, simple contrasts (first) were applied at the group level. For the control beliefs variable, the contrast between non-instruction and full-instruction groups was significant (p < 0.05). The confidence interval does not cross zero. Therefore, there is a good chance to observe group differences on this variable (95%) if applied to other samples from the same population.

These results indicated that a discrepancy between multivariate and univariate analyses occurred for the category of motivational beliefs for the variable control beliefs. It is possible that this is due to the fact that the groups involved in the examination differ along a combination of the dependent variables in this category;

therefore, to see how the dependent variables interact other statistical procedures are required, such as discriminant function analysis (Field, 2000).

A discriminant analysis of the category motivational beliefs (control beliefs, self-efficacy, and task value) at the group level was non-significant (Wilk’s Lambda, p

> 0.05). A confirmatory one-way ANOVA examining control beliefs across the groups was also non-significant (p > 0.05). However, using FTP as the grouping variable, discriminant analysis of the motivational beliefs variables are significant (Wilk’s Lambda, p < 0.05). This significance can be explained through the variables of self-efficacy and control beliefs, of which self-self-efficacy contributes the most to group separation (since it has the highest canonical variate correlation coefficient of the three dependent variables). Control beliefs is also important, but in relation to the other variables. Since it has a negative value that is close to -1, it confirms that any group differences are due to difference between variables.

9.1.2 Effects of FTP

All other significant differences were observed with the between-subjects factor of FTP degree.

9.1.2.1 Goal Orientation

Goal orientation indicated significant differences for mastery-approach (p <

0.05).

9.1.2.2 Motivational Beliefs

The only significant motivational beliefs variable using FTP as the grouping variable was task value (p < 0.05); however this level is almost non-significant. Post-hoc ANOVAs performed across both independent variables (group and FTP) revealed

Bonferroni test statistic for the Post Hoc ANOVAs, it is evident that students high in FTP have significantly different levels of self-efficacy than low-level FTP (p < 0.05).

The very low, almost non-significant, level of differences between students compared in terms of high/low FTP for the dependent variable task value were surprising24. According to theory on FTP and instrumentality, value in a task should be quite high for people who are identified as having a strong FTP recognizing the future value for the task. Further analysis was conducted to explore this result in greater detail.

As part of the demographic analysis of the participants, 5 items were included to assess reasons for participating in the course. Two of them related to the perceived future value (instrumentality) of the course – namely, usefulness and career importance.

Therefore, these aspects were examined through correlation of mean task value and reasons for course participation (reason-usefulness & reason-career).

Without controlling for type of instruction or level of FTP, results of the Pearson correlation analysis indicated a significant relationship between mean task value and reason-career (r = 0.21, N = 94, p < 0.05, one tail). The correlation analysis was repeated, this time controlling for level of FTP (binary – high/low). For students with low FTP (n = 35), no significant correlations were evident between mean task value and reason for course participation. However, for students with high FTP there was a significant correlation between mean task value and reason-career (r = 0.31, n = 59, p <

24 A reason for this might be in the item formulation for task value, since Pintrich and colleagues (1991) interpret task value in terms of the more general question of “what do I think about doing this task?” (e.g.

“I think I will be able to use what I learn in this course in other courses.”).

0.01, one tail). Having high FTP appears to increase the significance of career aspects for task value.

A final repetition of the correlation analysis controlling for instruction (see Table 9.2) revealed that students receiving future oriented instruction (full) reported levels for task value that correlated significantly with usefulness as a reason for course participation (r = 0.28, n = 44, p < 0.05, one tail). Modified instruction (non-future oriented) indicated similar results as before: levels reported for task value correlated significantly with reason-career (r = 0.34, n = 38, p < 0.05, one tail). For the non-instruction group (n = 12), no significant relationship between task value and reason for course participation were observed.

Table 9.2. Correlation matrix of task value and reason for course participation Time 1 Time 2 Time 3 Full Instruction

(n = 44) SD M 4.5 1.31 0.5 0.51 0.2 0.37

1. Task Value (max.=7) --.

2. Reason-Career (0=no; 1=yes) 0.11 --.

3. Reason-Usefulness (0=no; 1=yes) 0.28* 0.19 --.

Time 1 Time 2 Time 3 Modified Instruction

(n = 38) SD M 4.4 1.10 0.4 0.50 0.3 0.45

1. Task Value (max.=7) --.

2. Reason-Career (0=no; 1=yes) 0.34* --.

3. Reason-Usefulness (0=no; 1=yes) 0.94 -0.25 --.

Time 1 Time 2 Time 3 Non-Instruction

(n = 12) SD M 4.8 0.91 0.7 0.49 0.2 0.45

1. Task Value (max.=7) --.

2. Reason-Career (0=no; 1=yes) 0.13 --.

3. Reason-Usefulness (0=no; 1=yes) -0.19 0.00 --.

Note: * p < 0.05 (one-tailed)

9.1.2.3 SRL and Learning Strategies

SRL and learning strategies indicated many variables with significant differences, namely metacognitive-SRL (p < 0.05), time/study environment management (p < 0.01), and help-seeking (p < 0.01). Pairwise comparisons of these differences revealed that students with high levels of FTP reported higher levels for all of these variables, regardless of instructional group.

To follow-up on the MANOVA simple contrasts (last) were applied at the FTP level. For Metacognitive SRL, the contrast between low-FTP and high-FTP was significant (p < 0.05). Therefore, the likelihood of observing similar differences in Metacognitive SRL between students with high and low levels of FTP in other samples of the same population is fairly high (95%). This finding was supported in the Post Hoc analysis and was significant (Bonferroni, p < 0.05) indicating that high FTP also resulted in high metacognitive SRL (at least self-reported).

9.2 Investigation 2

Efforts to gain a better understanding of the data for this study continued into the second level of investigation which had three purposes:

ƒ To complete the main hypothesis examination of instructional effects by addressing the influence of future oriented instruction and FTP on academic achievement.

ƒ To explore the hypotheses regarding the relationship between perceived time perspective and the four other categories of variables included in the study (goal orientation, motivational beliefs, SRL, and achievement).

ƒ To explore the hypotheses regarding the stability or change in FTP and other dependent variables over the three phases of measurement.

Statistical operation of the first two purposes involved two series of regressions (an alpha level of .05 was used for all statistical tests). The first regression series tested the hypothesized model based on the theoretical background for this study in terms of how it predicts achievement. The second regression series examines the role of FTP in relationship to goal orientation, motivational beliefs and SRL.

The third purpose applied repeated measures statistics across the three time-frames examining within-subjects effects for each dependent variable and between-subjects effects for instructional group, gender and age.

Since the sample was reduced for this investigation (the non-instructional group was omitted due to non-participation in the third phase of measurement), an adapted overview of summary statistics is provided in Table 8.5.