The effects of peer instruction on students’ conceptual learning and motivation

The Results of FCI

Students’ conceptual understanding of Newtonian Mechanics using FCI (Hestenes et al., 1992) in both treatment and control groups was measured. To measure how well students performed after instruction relative to their performance before instruction, the Hake normalized gains (Hake, 1998) for each student was calculated. These sets of individual normalized gains for treatment and control groups were then compared for statistical significance.

Table 3 shows FCI scores before instruction (pretest) and after instruction (posttest), as well as the normalized gains (g) for students in traditional and PI course. Differences between the traditional and PI courses were considered significant for p values less than 0.05. No significant difference existed between PI and control sections before instruction; however, PI section achieved significantly greater normalized gains after instruction (p<0.01). Also it was found that the student normalized gains this group was higher found as g=0.59.

Table 3 Precourse (Spre) and Postcourse (Spost) FCI Results for PI and Traditional Courses

	n	S_pre	S_post	g
Treatment
Pretested	32	41.7	76.2	0.59
Not Pretested	30	-	75.3	-
Control
Pretested	31	42.1	60.8	0.32
Not Pretested	30	-	59.7	-
Difference		-0.4	15.4	0.27^a
^ap<0.01

Also, this test was analyzed statistically to determine whether the experimental intervention had affected students’ academic achievement as measured by the test. The 2x2 factorial analysis determined that the treatment groups performed significantly better on FCI than the control groups (F=5.03, p=0.012). No difference was found between those students who took FCI as a pretest and those who did not. This indicated that students who had taken FCI as a pretest had no significant advantage over those who had not (F=0.02; p=0.921). There was also no significant group by pretest interaction (F=0.34, p=0.543).

The Results of MSLQ

Factorial analyses were applied to determine whether the treatment had affected students’ motivation and self-efficacy. Analyses were therefore directed at the motivation (value and affective components) and self-efficacy scales within the motivation section of MSLQ as a measure of students’ motivation and self-efficacy. The means ± Standard Deviation “SD” for MSLQ scores are included in Table 4. Also the results of MSLQ categories were analyzed as follows:

Table 4 MSLQ Components

Source of Variation	n	Value Component	Affective Component	Self-Efficacy Component
Treatment
Pretested	32	4.88±0.97	3.62±1.43	5.13±1.12
Not pretested	30	4.79±1.03	4.13±1.31	5.02±1.08
Control
Pretested	31	4.73±0.79	3.81±1.31	4.61±1.09
Not pretested	30	4.61±1.01	3.95±1.38	4.45±1.15

Value Component Results

The value component of MSLQ motivation scale measures students’ interest and goal orientation and the value of the course. Higher means indicate more interest, value, and positive goal orientation in the course and serve as a measure of students’ motivation. The factorial analysis revealed no significant differences between the treatment and control groups with regard to students’ scores on the value component of the motivation scale of MSLQ (F=1.26, p=0.382), even though the means were higher in the treatment groups (Table 5). No differences were noted between the pretested groups (F=0.00, p=0.847). Likewise, there was no significant pretest by treatment interaction (F=1.07, p=0.311).

Table 5 Value Component of MSLQ (n=123)

Source of Variation	SS	df	MS	F	p
Main effects Within (error)	124.56	119	1.04
Treatment vs. control group	1.32	1	1.32	1.26	0.382
Pretest	0.00	1	0.00	0.00	0.847
Group by pretest	1.12	1	1.12	1.07	0.311
Between (model)	2.44	3	0.81	0.77	0.591
Total	127	122	1.04
SS: sum of squares; df: degrees of freedom; MS: mean squares.

Affective Component Results

The affective component of MSLQ motivation scale measures how much students worry about tests and how often they have distracting thoughts when they take an exam. Higher means indicate more anxiety in testing situations, thus measuring the affective component of motivation. The factorial analysis revealed no significant differences between the treatment and control groups with regard to students’ scores on the affective component of the motivation scale of MSLQ (F=0.18, p=0.664; Table 6). No differences were noted between the pretested groups (F=1.97, p=0.152), and there was no significant pretest by treatment interaction (F=0.70, p=0.433).

Table 6 Affective Component of MSLQ (n=123)

Source of Variation	SS	df	MS	F	p
Main effects Within (error)	235.81	119	1.98
Treatment vs. control group	0.36	1	0.36	0.18	0.664
Pretest	3.92	1	3.92	1.97	0.152
Group by pretest	1.39	1	1.39	0.70	0.433
Between (model)	5.67	3	1.89	0.95	0.445
Total	241.48	122	1.97

Self-efficacy Component Results

The self-efficacy component of MSLQ measures students’ expectancy of success, their perceptions of self-confidence in understanding the course content, and their control over those beliefs. Higher means indicate the better students believe they will do in the course and be able to master the course material. The factorial analysis revealed a significant difference between the treatment groups and the control groups with regard to their self-efficacy on the self-efficacy section of MSLQ (F=4.37, p=0.041; Table 7). No significant differences were found between those who were pretested and those who were not (F=0.37, p=0.563), and there was no significant pretest by treatment interaction (F=0.02, p=0.879).

Table 7 Self-Efficacy Component of MSLQ (n=123)

Source of Variation	SS	df	MS	F	p
Main effects Within (error)	152.18	119	1.27
Treatment vs. control group	5.55	1	5.55	4.37	0.041
Pretest	0.47	1	0.47	0.37	0.563
Group by pretest	0.03	1	0.03	0.02	0.879
Between (model)	6.05	3	2.01	1.58	0.214
Total	158.23	122	1.29