# Not sure if i chose the correct statistical test to analyze data

#### hugstari

##### New Member
I'm trying to help a friend out with the statistics parts of her bachelors paper. Right now she is still the the planing stages and has not gathered data yet. She is trying to figure out the effects of peer-support counseling for cancer patients and their family. Her problem (and mine) is figuring out the correct statistical test to use for when she has gathered the data.

She is doing a questionnaire survey and the first few questions are just your basic sex\age\education type questions. After that there are a couple questions about how they heard about the support counseling, how many times they received support, over how long a period the counseling lasted and finally how many times the counselor initiated contact with the person seeking support. After those questions there are about 30 questions, most of them are on a five point likert scale but a couple provide the sixth option, which is "none at all". These questions all have something to do with the peer-support counseling and how it affect the one seeking support in a variety of ways. Here is an example of the type of questions that are asked with the five point answer option and the six point answer option:

Five point scale answer option: "How satisfied or unsatisfied were you with the length of support concealing?" "How much or little information do you think the peer-counseling provided you with for your disease?"

Six point scale: "How satisfied or unsatisfied were you with the peer-counseling you received from the organization?" "If you experienced social isolation in the wake of your diagnoses, how much or how little did the peer-counseling help you break free from the isolation?"

She still has not deiced how she is going to analyze the answers but thinks she might do it based on sex, age and frequency of counseling sought. I told her that it might be better combining some of the questions into smaller groups and calculating a cumulative score for those groups based on the type of questions. Kind of like a factor analysis but without a actually doing a factor analysis because i don't think she has a large enough sample to meet the requirements for the FA. I told her that I thought the right way to go forward would be to put the data into a contingency table and doing cross tabulations and a chi square test. Then she could also do some visual analysis with bar charts and the like. Then finally present some descriptive statistics like correlations and confidence intervals.

But the problem remains, I am still not sure that this is the best way to analysis the data from the questionnaire. Perhaps i've sent her down the wrong path and there is a better or easier way to analysis the data. Any help in the right direction or confirmation that she (we) are on the right path would be great. I'll be happy to provide any further information if i'm missing something from my description, odds are i am