• Tidak ada hasil yang ditemukan

Analysis of variance on knowledge of HIV/AIDS transmission

9. Money for upkeep on campus: 1 for “Not enough”, 2 for “Just enough”, 3 for

5.12 Socio-demographic characteristics and knowledge about HIV/AIDS

5.12.3 Analysis of variance on knowledge of HIV/AIDS transmission

One-way ANOVA and non-parametric K tests for several independent samples were done.

While references were made to data from One-way ANOVA, only data for the non- parametric tests are presented in Tables 5.16 a.

ALL: Statistical significant relationships are observed for ‘How old were you in your first year in University?’ (F-ratio = 3.094, df = 2, Sig. = 0.046) and ‘How old are you now?’ (F-ratio = 3.169, df = 2, Sig. = 0.042) from One-way ANOVA (parametric) tests on relationships between the first 9 items in Section A and the categorised total scores for Section G, thus the null hypothesis is rejected in these two cases and accepted for the other relationships;

that is, only these two variables would have any significant contribution to knowledge of HIV transmission in the whole respondents.

However, from the non-parametric data four variables, including the two variables above, have significant statistical relationships: ‘How old are you in your first year in the University’,

How old are you now?’, ’What is your marital status?’, and ‘How many children do you have now?’. Thus it could be concluded that the null hypothesis is rejected in these four variables in the entire population and thus have significant effect on knowledge of HIV transmission.

UNAD: The significance is >0.05 in all variables, in both One-way ANOVA and non- parametric tests, leading to the acceptance of the null hypothesis for UNAD. This implies that all the variables considered in this analysis, curiously, did not have significant effect on the knowledge of transmission of HIV/AIDS among the respondents in this institution.

UNIZULU: It is only in ‘How old are you now?’ (F-ratio = 3.281, df = 2, Sig. 0.038), from the One-way ANOVA, that significant relationship was observed, which reflects the rejection of the null hypothesis for this variable only. In all other variables significance was >0.05 leading to the acceptance of the null hypothesis. However, the same set that showed significant statistical relationship for the non-parametric tests in the entire population were

also significant in UNIZULU. Other variables had no statistical significance and therefore did not have effect on knowledge of transmission.

Thus it could be concluded that the null hypothesis was rejected in the four variables identified under ALL and UNIZULU but not for UNAD. The implication is that age, marital status, and number of children have significant effects on knowledge of HIV/AIDS transmission, but all other variables did not for UNIZULU’s respondents while all variables tested did not have effects on the UNAD’s respondents’ knowledge on transmission.

5.12.4 Analysis of variance on knowledge of protection against HIV/AIDS infection The data for non-parametric tests are presented in Table 5.16 b. From One-way ANOVA, statistically significant relationships were found between ‘Which of these best describes where you grew up as a young boy or girl’ in ALL (F-ratio = 3.474, df = 2, Sig. 0.031), UNAD (F-ratio = 6.614, df = 2, Sig. 0.001), and UNIZULU (F-ratio = 4.398, df = 2, Sig. 0.013), and

How old were you in your first year in University’ in ALL (F-ratio = 3.160, df = 2, Sig.= 0.043) but not in UNAD and UNIZULU and knowledge of protection against HIV/AIDS infection. The null hypothesis may therefore be rejected for each of these relationships and accepted for the other variables.

Corresponding analysis using the non-parametric tests reaffirm significant statistical relationships for ‘How old were you in your first year in University’ in UNIZULU and ALL,

How many children do you have?’ for UNIZUL and ALL, ‘What is your religion?’ for UNAD and ‘Which of these best describes where you grew up as a young boy or girl’ in ALL, UNAD and UNIZULU as well as an additional variable, ‘How would you describe your family’s household resources?’ for ALL.

Table 5.16 a: Results from ANOVA using non-parametric test on selected characteristics and knowledge of transmission Section G (Only those with significant effects are listed)

Independent variables Knowledge of HIV/AIDS transmission

UNAD UNIZULU ALL

How old were you in your first year in University? Chi-Square df

Asymp. Sig.

.661 2 .719

6.030 2 .049

7.570 2 .023

How old are you now? Chi-Square

df

Asymp. Sig.

.446 2 .800

7.289 2 .026

7.596 2 .022 What is your marital status? Chi-Square

df

Asymp. Sig.

.888 2 .641

7.159 2 .028

7.000 2 .030 How many children do you have? Chi-Square

df

Asymp. Sig.

.838 2 .658

7.328 2 .026

7.132 2 .028

Evidently age, family resources and the environment where respondents grew up as youths, number of children and family resources appear to have significant effects on the respondents’ perception of prevention against HIV infection; other variables do not. Religion also does for UNAD only.

Non-parametric t-test was run for the equality of means of total scores for Sections G and H using the Mann-Whitney t-test. The results are tabulated in Table 5.17. With a significance of 0.000 in each case the null hypothesis is rejected, thus the mean scores from the two institutions are not equal.

Table 5.16 b: Results from ANOVA obtained from non-parametric tests on selected characteristics and knowledge of protection against HIV/AIDS infection (Section H)

Independent variables Knowledge of protection against HIV/AIDS

UNAD UNIZULU ALL

How old were you in your first year in University?

Chi-Square df

Asymp. Sig.

2.755 2 .252

6.265 2 .044

9.944 2 .007 What is your marital status? Chi-Square

df

Asymp. Sig.

2.363 2 .307

2.372 2 .305

3.692 2 .158 How many children do you have? Chi-Square

df

Asymp. Sig.

2.174 2 .337

8.163 2 .017

9.727 2 .008 What is your religion? Chi-Square

df

Asymp. Sig.

7.725 2 .021

.180 2 .914

3.008 2 .222 Which of these best describes where

you grew up as a young boy or girl?

Chi-Square df

Asymp. Sig.

11.394 2 .003

6.096 2 .047

7.327 2 .026 How would you describe your

family’s household resources?

Chi-Square df

Asymp. Sig.

4.908 2 .086

.986 2 .611

5.974 2 .050

Table 5.17: Mann-Whitney t-test for equality of means (Total scores)

Institution N

Mean Rank

Sum of Ranks

Mann- Whitney U

Wilcoxon W Z Asymp. Sig.

(2-tailed) GTOTAL UNIZULU 856 783.54 670710.50 213109.500 395819.500 -5.725 .000

UNAD 604 655.33 395819.50 Total 1460

HTOTAL UNIZULU 856 767.99 657401.00 226419.000 409129.000 -4.047 .000 UNAD 604 677.37 409129.00

Total 1460

5.13 Knowledge of preventive measures