I am conducting a survey to compare information gathered over the internet with information coming from face-to-face communication.Needed help with the statistical analysis after data entered in spss.
You will need to provide more information in order for anyone to give you any real advice.
What is your research question/primary hypothesis? You need to define this both in a practical sense AND in a statistical sense. For example, that face-to-face communication improves some measure X, with some sort of a priori idea of how big an increase in X is considered an improvement. Since you didn't specify what kind of information you gathered or why you gathered it, it is impossible to help you formulate this sensibly.
What is your study design? What are you trying to compare? Performance in some measure between two groups? In one group across time?
What kind of data are you gathering?
etc.
What has the literature shown you regarding similar designs? You may also want to obtain a book on spss to aid in the choice of statistic to use, what is the appropriate statistic and how to interpret it. Also, may be helpful to talk with a professor of statistics. Good luck!
In case of numerical data like height, weight, haemoglobin, income, sales, you can use t-test or normal test for comparison of means of any two groups of interest. In case you are dealing with percentages or proportions, use normal proportion test. In case of qualitative data like preference to music where you are able to categorize the data into groups, you can use chi-square test. Let us say, I wish to see whether possession of TV, Fridge or car varies by different socioeconomic status then essentially I am dealing with frequencies. In such case, use chi-square test.
As Ramnath Takiar points out, in case of data like height, weight, correlation, etc, there are several tests that you can use to access the similarity between the two data sets. The test you will use depends on how large the data set is, it's distribution and several other factors. Two of the main test's that statisticians use to compare data set's are:
You can find link's to two of these below. Since you have not mentioned anything about your data, here are several other options that you can use:
https://en.wikipedia.org/wiki/Student's_t-test
https://en.wikipedia.org/wiki/Kolmogorov%E2%80%93Smirnov_test
https://en.wikipedia.org/wiki/Kruskal%E2%80%93Wallis_one-way_analysis_of_variance
https://en.wikipedia.org/wiki/Jaccard_index
https://en.wikipedia.org/wiki/Mann%E2%80%93Whitney_U_test
https://en.m.wikipedia.org/wiki/S%C3%B8rensen%E2%80%93Dice_coefficient
There are too many questions that need an answer before we can help you. Jochen's suggestion to contact a local statistician is the best start. If you have more specific questions afterwards, we can try to help.
Before you see the statistician, please have a clear hypothesis. Also, you need to see the statistician to help determine sample size. Do not wait until you have collected all your data. Too many times a good experiment is ruined by insufficient data. Finally, try very hard to avoid missing data.
- Badges
- Science topic
- Similar topics
- Biological Science
- Histology
- Fixatives