Methods of Communication Research and Statistics

Do older people spend more time watching television than younger people? Or do younger people actually watch more? Is there any difference between watching television on work days and at weekends? Use the following database: McoData2.sav. Use the variables you created in exercise 6.1. again for this assignment. If you have already done this exercise elsewhere, you will have to create the variables again: 'minutes watching television on work days' (v5hour and v5min) and 'minutes watching television at weekends' (v6hour and v6min).

a. Generate a scatter plot with the independent variable being 'age' (v3) and the dependent variable being 'minutes watching television on work days'. Is there a weak, moderate, or strong association?

b. Generate a scatter plot, with the independent variable being 'age' (v3) and the dependent variable being 'minutes watching television at weekends'. Is there a weak, moderate, or strong association?

c. Calculate the Pearson's correlation coefficient for the 'age' and 'minutes watching television on work days' variables and provide your interpretation.

.01, no association between age and minutes spend watching television on workdays

d. Calculate the Pearson's correlation coefficient for the 'age' and 'minutes watching television at weekends' variables and provide your interpretation.

59, strong positive association, the older someone is, the more television one watches on a weekend day.

There is a significantly strong association between amount of time spent watching television on workdays and amount of time spent watching television on weekend days: r = .59, p < .001.

e. Conduct a regression analysis, with the independent variable being 'age' and the dependent variable being 'minutes watching television on work days'. What is the literal meaning of the value of the intercept?

a person of 0 years old watches 140.17 minutes of television

f. What is the value of the unstandardised regression coefficient of the predictor age?

g. What does this value of the unstandardised regression coefficient mean? Mention the variables about which you are making a statement in your answer.

when age increases by one unit, the number of minutes spent watching television on workdays increases by 0.09. However, the t test statistic is not significant, meaning that the unstandardized regression coefficient in the population may be zero. That means that age is not a significant predictor of minutes spent watching television on work days.

The value indicates that there is no effect on minutes spent watching television on workdays when age increases, because it is very low (almost zero).

Did you notice that the standardised regression coefficient corresponds to the product-moment correlation coefficient?

j. Does age have a significant effect on the number of minutes spend watching television on workdays? Name the value of the statistic and the significance, and what it means for the null hypothesis.

no, the result is not significant. t = 0.24, p = .808. We therefore retain the null hypothesis; age has no effect on time spent watching television on workdays.

k. Provide your interpretation of the 95% confidence interval for the 'age' variable.

there is 95% certainty that the unstandardised regression coefficient in the population lies between -0.64 and 0.82

l. On the basis of the confidence interval, can you say something about the significant of the test?

not significant, the regression coefficient in the population can be both negative and positive. The value zero falls in the confidence interval, which is an indication that there is no significant result. It is possible that the regression coefficient in the population is zero.

0.0% of the variance in age explains the variance in the number of minutes spent watching television during a workday

o. Conduct another regression analysis, but now with 'minutes watching television on weekend days' as the dependent variable. Report all the results in APA format.

In this sample, the model predicts that if someone is zero years old, they spend 182.69 minutes watching television on weekend days and that with each one-year increase, the number of minutes decreases by 0.06. However, this is not significant: b* = 0.01, t = -0.13, p = .894, 95% CI [-0.87, 0.76]. The variance in age explains 0.0% of the variance in time spent watching television on weekend days.

SPSS exercise 6.2