1. Describe, briefly, your observational study or experiment and state your hypotheses. What do you think is the correlation of the variables and why.
2. Select a well-designed sample of at least 30 units. (Describe the population to be studied and the sampling frame). Two variables should be measured with a potential for correlation; if you measure more than two variables, you must correlate two at a time. Pay particular attention for any lurking variables. If you are doing an experiment make sure there is a control group.
3.Care must be taken in the collection of the data. Avoid sampling and non-sampling errors. If working with subjects (people, that is), be careful with the wording of your problems. If you are using secondary data, don't forget to document the sources.
4.Give descriptive statistics for each single variable. At a minimum you should include the mean, median, mode, standard deviation, and five-number summary. A frequency table as well as a histogram should be included for each numerical variable.
5.Correlation analysis should include a scatterplot and the line of best fit (linear regression). Give the correlation coefficient as well as .
6.Draw your conclusions or inferences. describe the strength of the association. Be careful not to conclude more than your correlation analysis allows; that is, do not make any statements not supported by your data. How confident are you about your inferences?