Question

You, recently participated in an experiment to test the effectiveness of saturation patrol on residential burglaries. Prior to the experiment there were two neighborhoods (both about the same size) in your city with an equal number of residential burglaries. Burglaries in these neighborhoods were inordinately high. Prior to the experiment both neighborhoods had the same number of patrol officers assigned to them across all shifts. The captain assigned twice the normal number of patrol officers to work in Neighborhood A (i.e. saturation patrol) in an effort to deter future burglaries. The captain did not change the normal number of patrol officers to work in Neighborhood B. At the end of six weeks the number of residential burglaries decreased in Neighborhood A but remained the same in Neighborhood B. The chief wants to know if the difference in residential burglaries between these two neighborhoods is statistically significant, i.e. not due to change. Answer the following questions. 1. What is the independent variable in this experiment? 2. At what level (nominal, ordinal or scale) is the independent variable measured? 3. What is the dependent variable in this experiment? 4. At what level (nominal, ordinal or scale) is the dependent variable measured? 5. What is the null hypothesis for this experiment? 6. What is the alternative (research) hypothesis for this experiment? 7. What type of hypothesis (difference or association) is the alternative (research) hypothesis? 8. Which of the statistical techniques that you have learned so far (Chi-Square or t-test) would be the most appropriate to analyze the data in this experiment?

Answer 1

1. Independent variable : saturation patrol

2. Level of measurement : nominal

3. Dependent variable : number of residential burglaries

4. Level of measurement : scale

5. Null hypothesis : There is no significant difference in residential burglaries between two neighborhoods

6. Alternative hypothesis : There is significant difference in residential burglaries between two neighborhoods

7. Type : Difference

8. Test : t test

Note :We are interested to know whether saturation patrol decreases number of burglaries. So independent variable :saturation patrol and dependent variable : number of residential burglaries . The independent variable , saturation patrol takes two values 'yes' or 'no' . If number of patrol officer is normal , means saturation patrol 'no' . If its doubled , saturation patrol 'yes' . Thus independent variable takes nominal measurement . The dependent variable can be measured , so it is scale variable. As the dependent variable is scale , and the chief wanted to know if the difference is significant , we use t test for difference in means . We use chi square test when the dependent variable is nominal .