Chapter 7 Evaluation of the SETSIS
7.2 Association of the SETSIS and science teaching practice
7.2.1 Professional Teaching Practice Assessment
In the BoTP curriculum, Professional Practice comprises formal modules for PSTs practicing their teaching in real situation. These modules are conducted in real primary science classes during the final four semesters of the BoTP programme. These modules are part of the ITE curriculum used to assess in-site practices. PSTs need to undergo three phases of modules of Practicum in semesters 5, 6 and 7, and a module of Internship in semester 8 at designated schools.
During the practicum, a PST needs to teach for at least eight periods in science classes every week. Teacher trainers and guidance teachers from the schools observe the teaching sessions in real classes and conduct supervision of nine times in every phase. Joint assessments on teaching performances are based on
y = 1.3877x - 7.9666 R² = 0.1331 15 17 19 21 23 25 27 18 19 20 21 22 23 24 25 Ob se rv ed T ISP score
TISP prediction score
Model prediction
institutional criteria assessment at the end of every phase (Institute of Teacher Education, 2016).
During the Internships, PSTs need to teach about four periods of science subject classes, plan and implement activities related to science learning for the schools. PSTs receive guidance from their mentor (e.g. existing in-teachers) and from their teacher trainers during their time in the schools. A joint formative assessment from the mentor and the trainers was conducted based on institutional criteria assessment at the end of the module.
Figure 27 Frequency of score of Professional Practices for Practicum phase II (above) and Internship (below)
Figure 27 shows the frequency of score of assessments of Professional Practices received from ITE’s Department of Exam and Senate. The results of this section are bound to caveat from the usefulness of data received from the institutions. The following data present two out of the four formal modules conducted at ITE. The histogram of the Practicum II module assessment shows 49 scores distributed in the range 72 to 100. The histogram of Internship module assessment shows 64 scores distributed in the range 69 to 100.
Table 67lists the descriptive table of assessment scores in Practicum II and
Internship. The mean scores reflect that both modules have approximately the same performance with an Internship mean score at 87.09 (SD=7.14) and Practicum II mean score just slightly below at 86.63 (SD=7.18). The scores of the Internship
module have a slightly wider range with significantly normal distribution compared to the Practicum II module, which were not significantly normal.
Table 67 Descriptive table for Practicum II and Internship
Due to the following analysis of multiple regression, the scores for Practicum II and Internship were combined and label as Professional Practice. Figure 28 shows the histogram of the combined scores of the two assessments called Professional Practice. Visual inspection shows good distribution of the sample in the assessment score.
Figure 28 Performance in Professional Practice
The combination of 113 samples from the two module assessments were used for this analysis. Descriptive analysis in Table 68 shows the result of mean performance in Professional Practice within the range of the SD of the two mean scores from Table 67. A normality test shows a significant result (p<0.01). The result concludes that the combination data was appropriate to represent the performance of the two modules using Professional Practice.
Statistic df Sig. Practicum II 49 86.63 7.18 1.03 72.00 100.00 0.97 49.00 0.21 Internship 64 87.09 7.14 0.89 69.00 100.00 0.95 64.00 0.01 Shapiro-Wilk N Mean Std. Error SD Minimum score Maximum score
Table 68 Descriptive analysis of Professional Practice score
7.2.2 Correlation between the SETSIS and teaching practice assessment
The SETSIS measurement model were developed theoretically using the concept of self-efficacy believe to predict capability in performing the task of teaching science using science inquiry skills. Using the concept in the model, this section tests the hypothesis that the SETSIS measure is capable in predicting the performance of respondents in the practice of teaching science (i.e. the assessment score of Professional Practice).
Correlation and multiple regression analyses were conducted to examine the relationship between PSTs’ assessment scores of Professional Practice and potential predictors from the SETSIS measure. The results are used to discuss potential utilisation of the SETSIS model in the area of science teaching practice. Bivariate correlation between performances in professional teaching practice with the three theoretical factors model of the SETSIS are listed in Table 69. Pearson
correlations show non-significant correlation between the practice and any
components of the SETSIS. However, it can be seen that each of the practice scores correlate very weakly with the predictors. All predictors were positively correlated except PTE, indicating higher traits in predictors tending to have higher performance in practice except for the PTE trait.
Table 69 Correlation of Professional Practice with the SETSIS model
Two multiple regression models were tested. Model 1 has three possible predictors (i.e. KE, PTE and OBE) and Model 2 has two possible predictors (i.e. KE and PTE).
Statistic
df
Sig.
Teaching Practicum 113.00 86.89
7.13
69.00
100.00
0.96
113.00
0.00
Shapiro-Wilk
N
Mean
SD
Minimum
score
Maximum
score
KE PTE OBE Professional Practice 0.104 -0.018 0.009 Professional PracticeChecking on residual plots of both models show random patterns that indicate that the data can be used with the regression models. Table 70 lists the results of regression model summary for Models 1 and 2. Results show both models have weak predictor correlation indicating that both models explain only 5 percent of the variability of the performance data around its mean.
Table 70 Results of regression model summary to infer Professional Practice
Model R Adjusted R square
Standard
Error F-test Significant Model 1 0.225 0.015 7.077 1.434 0.228 Model 2 0.226 0.034 7.010 2.969 0.055
Results of adjusted R square in Table 70 increase in Model 2 instead of Model 1. The result indicates that the model predicting better without OBE. F-test shows that Model 1 does not significantly work, F (3,109) = 1.43, p>0.05. Model 2 shows a better working model with F-test result is just slightly over significant F (2,110) = 2.97, p=0.055.