Effectiveness of hands-on and minds-on activities on students’ achievement and attitudes towards physics

The data obtained from the pretest and posttest achievement and attitude scores of all students were analyzed both descriptively and inferentially. Descriptive results related to the pretest and posttest achievement and attitude scores of all students are given in Table 2. The experimental group gained a mean increase of 2.16 points while the control group gained 1.09 points from pretest to posttest on the PACT. The mean increase on the PATS was 5.47 points for the experimental group while the mean increase was 2.59 points for the control group from pretest to posttest. Furthermore, the standard deviation values were relatively stable from pretest to posttest on the physics achievement and physics attitude scores.

Table 2: Basic descriptive statistics related to PACT and PATS scores

	Experimental Group		Control Group
	Pretest	Posttest	Pretest	Posttest
Scores on Physics Achievement Test *
N	70	70	60	60
Mean	10.64	12.80	8.99	10.08
Standard Deviation	2.84	3.31	2.04	3.13
Scores on Physics Attitude Scale **
Mean	80.79	86.26	78.51	81.10
Standard Deviation	13.05	14.25	12.66	15.60

* Possible minimum and maximum PACT scores are 0 and 21, respectively

** Possible minimum and maximum PATS scores are 24 and 210, respectively

To statistically equalize the differences among the experimental and control groups, six independent variables; students’ age, gender, pre achievement and pre attitude scores, previous physics course grades, and previous cumulative grade point averages were planned to be used as covariates. All pre-determined independent variables have been correlated with the two dependent variables. All independent variables -except students’ age- had significant correlations with at least one of the dependent variables so were determined as covariates for the following inferential analyses.

Before using the statistical model of multivariate analysis of covariance (MANCOVA), normality, homogeneity of regression, multicollinearity, equality of variances, and independency of observations assumptions were checked. After validation of these assumptions, the MANCOVA model was conducted. Significant differences were found among hands-on/minds-on instruction and traditional methods on the collective dependent variables of the post achievement (PSTACH) and post attitude scores (PSTATT), F(2,120) = 3.396; p < .05. In order to test the effect of the methods of teaching on each dependent variable, a univariate analysis of covariance (ANCOVA) was conducted as follow-up test to the MANCOVA. The effect of those activities on the PSTACH, F(1,121) = 5.781; p = .018 was significant. Therefore, the results suggest that students instructed by hands-on/minds-on activities had a higher achievement in physics than the students taught by traditional method. However, the results on the PSTATT, F(1,121) = 1.368; p = .244 were not significant. Moreover, the statistical analyses failed to show any significant differences between the means of the students’ attitude subcategories (enjoyment, self efficacy, importance of physics, achievement-motivation, and interest related behavior) towards simple electric circuits.

A step-down analysis was also used as a follow-up analysis subsequent to the MANCOVA. This analysis was used to investigate the unique importance of dependent variable, the PSTACH, which was found as significant in the ANCOVA. When the students’ physics achievement scores were analyzed with physics attitude scores acting as an additional covariate, the effect of those activities was still significant, F(1,120) = 5.375; p = .022. This indicates that the effect of the hands-on/minds-on activities on students’ physics achievement after accounting its effect on physics attitude was also significant. That is, physics achievement was also significantly and uniquely affected by the hands-on/minds-on activities after their effects on physics attitude.

Throughout the study, both the experimental and control groups have been observed for the purpose of the treatment verification. Mann-Whitney U test and independent t-test were conducted for those observation results. Results of these tests were significant which indicate that the treatments in the experimental and control groups were significantly different. Moreover, the means and the standard deviations of each item for both groups were presented in Table 3. It indicates that for the experimental group, means of positive items (hands-on activity criteria) for the experimental group were drastically greater than the means of the control group while the means of negative items (8 and 9) for the experimental group were drastically lower than the means of the control group. These results verified that lessons in the experimental group were implemented according to the hands-on activity criteria and those in the traditional group were implemented according to the traditional method. Therefore, treatment verification was supported.

Table 3: Basic descriptive statistics related to items of the observation checklist

Item Number	Experimental Group		Control Group
	Mean	Std. Dev.	Mean	Std. Dev.
1	3.00	-	0.00	-
2	3.00	0.58	0.00	-
3	2.71	0.49	1.86	0.69
4	2.71	0.75	0.00	-
5	2.57	0.53	1.43	0.53
6	3.29	0.49	1.43	0.53
7	3.00	0.82	1.43	0.53
8	1.00	-	3.57	0.53
9	1.14	0.38	3.71	0.49
10	3.57	0.53	0.00	-
11	2.14	0.38	0.00	-
12	2.57	0.53	0.00	-