Inferential Statistics: Movies and Behavior

Depending on the kinds of themes they expose, movies have different impacts on viewers. Action movies tend to promote violent behaviors when compared to comedies and dramas. The establishment of the relationship between the nature of films and the occurrence of violent acts among children is necessary. In essence, the analysis of the effects of movies on violent behavior provides critical information for appropriate parenting. The study of movies from 1937 to 1999 using inferential statistics would indicate whether they have significant effects on children. Inferential statistics aid in decision-making since they determine if the outcomes of data analysis are valid. Typically, alpha levels of 0.1, 0.05, and 0.90 are used as thresholds of rejecting or accepting a null hypothesis of a given study (Field, 2017). In this case, the alpha level of 0.05 will be used to determine if the exposure period to movies has unique and significant effects on violent behavior in children. Based on the hypothesis that movies have different effects, this analysis aims to evaluate and identify a period of movies that influence violent behaviors in children.

Independent T-Test

The independent t-test was used to test the null hypothesis that the two periods of movies do not have statistically significant differences in the mean number of injuries in children. By using SPSS, the independent t-test was employed to assess whether movies created in 1937-1980 or 1981-1999 have different effects on violent behavior among children. Before analysis, the dependent variable (the number of injuries) and the independent variable (the period of movies) were checked to ensure that they comply with major assumptions of the independent t-test. The dependent variable should exist on an interval or ratio scale, follow the normal distribution, possess homogenous variance, and be free of significant outliers (Field, 2017). In this test, the number of injuries meets these assumptions and generates robust results. Regarding the independent variable, the t-test assumption requires it to have two independent groups (Pallant, 2016). The independent variable conforms to this assumption as it categorizes movies into two groups, one period of 1937-1980 and another period of 1981-1999.

According to the results of the t-test (Table 1), the exposure period of 1937-1980 caused a statistically significantly lower number of injuries than the exposure period of 1981-1999 (M = 2.12), t(72) = -3.10, p = 0.003. These results suggest that recent movies created between 1981 and 1999 are more violent when compared to the ones produced in 1937-1980.

Table 1.

Independent Sample t-Test Results

		Levenes Test for Equality of Variances		t-test for Equality of Means
		F	Sig.	t	df	Sig. (2-tailed)	Mean Difference	Std. Error Difference	95% Confidence Interval of the Difference
									Lower	Upper
Injuries	Equal variances assumed	9.439	.003	-3.100	72	.003	-1.379	.445	-2.265	-.492
	Equal variances not assumed			-3.914	71.100	.000	-1.379	.352	-2.081	-.676

One-Way ANOVA

The examination of the dependent variable of the number of injuries shows that it comply with the assumptions of homogeneity of variance and interval scale (Gravetter & Wallnau, 2017). Levenes test indicates that there is equality of variances in the distribution of injuries in all the groups. The number of injuries exists on an interval scale to allow its analysis as the dependent variable in one-way ANOVA. Additionally, the scatter and Q-Q plots reveal that the number of injuries follows the normal distribution. Since the independent scale of the period of movies has three groups, it observes the assumption of required categories in ANOVA.

ANOVA was used to test the null hypothesis that the mean number of injuries in three periods of movies do not have statistically significant differences. Outcomes of ANOVA (Table 2) depict that the mean number of injuries were lowest in 1937-1950 (M = 1.00, SD = 1.044), moderate in 1961-1980 (M = 1.59, SD = 1.992), and highest in 1981-1999 (M = 1.95, SD = 1.974). These outcomes suggest that movies get violent with time because of the increasing number of injuries in children. Furthermore, standard deviations reveal that variations in the number of injuries increase as their means grow. However, ANOVA shows that the apparent increases in the number of injuries with the exposure periods are statistically insignificant, F(2,71) = 4.431, p = 0.284. The inference here is that the obvious differences in means of the number of injuries in the three periods are not enough to offer valid evidence to support the alternative hypothesis.

Table 2

ANOVA Results of Injuries in Three Periods

	N	Mean	Std. Deviation	Std. Error	95% Confidence Interval for Mean		Minimum	Maximum
	N	Mean	Std. Deviation	Std. Error	Lower Bound	Upper Bound	Minimum	Maximum
Period 1937-1960	12	1.00	1.044	.302	.34	1.66	0	3
Period 1961-1989	22	1.59	1.992	.425	.71	2.47	0	6
Period 1990-1999	40	1.95	1.974	.312	1.32	2.58	0	9
Total	74	1.69	1.872	.218	1.26	2.12	0	9

Descriptive statistics show that the period of 1937-1960 has the number of injuries ranging from 0 to 3 [CI: 0.34-1.66], while those of the period of 1961-1980 varies from 0 to 6 [CI 0.71-2.47]. The period of 1990-1999 has the highest mean number of injuries in which values range from 0 to 9 [CI: 126-2.12].