Propose a research hypothesis that can be tested using any of the five statistics discussed in this module. State the null hypothesis
A hypothesis is a question which a researcher looks to answer by showing the relationship between two or more phenomena he or she chooses. Null hypothesis is the opposite of the question a researcher is looking to study. My hypothesis is “Do people engage in reckless driving simply because they have car insurance?” Null hypothesis would be people engage in reckless driving.
Identify the variables in your proposed hypothesis
Variables are the logical classification of attributes into independent or dependent variables. Where variables cannot be manipulated, predictor valuable and outcome variable is used. Predictor variable in my hypothesis would be car insurance while the outcome variable would be reckless driving.
Describe the levels of measurement in each variable in your hypothesis
There are four levels of measurement which include nominal, interval, ratio and ordinal measurement. To measure the variable, I will randomly select a group of people and ask them whether they think drivers drive recklessly because they have car insurance. I will then use nominal measurement to quantify this data. For the car insurance I will use ordinal measurements which depict an orderly attribute but there’s no clear quantifiable measure between reckless driving and car insurance. Answers I might get include highly likely, less likely and that’s why I shall use ordinal measurement.
Determine what test statistic you would choose to test your proposed hypothesis. Briefly justify your choice
There are two ways to test hypothesis which are Correlation research and experimental research. Correlation research includes observing what naturally happens while experimental research involves manipulating the variables in order to get the desired outcome. In testing my hypothesis, I would use correlation research by observing the responses I receive from conducting interviews where I ask my interviewees whether they think drivers engage in reckless driving because they have car insurance. Correlation research will allow me to check the responses I receive and get an outcome without manipulating any variable.