These are dependent measures. So you cannot do a oneway analysis
of variance, instead you'd need a repeated-measures analysis of
variance. The results of such an analysis are influenced by the degree
of correlation between answers to these questions. You can calculate
the correlations only if you have the raw data.
Moreover, can we assume that the answers to all 5 questions were
given on the same scale? The results for the last question look
I did many searches and understand the meaning of these measurements but what will the reader benefit from knowing that variance of any question is small and so close to mean -I am not sure If I understand it right or not!-. For example, Q2 is about age, there are 5 age groups, the mean is 3,24, so can I write that the mean shows that the majority of respondents are within 35-44 age group? please correct me if I am wrong and what about the rest of measures?
I agree that there is dependence in the data based on what you've told us. I do think generally it would only increase power to model that dependence though.
It is entirely possible to do an ANOVA assuming independent observations with what you have. I'm not sure it's worth it though and it sounds like you're just more concerned with how to present the data and I don't think doing additional tests that weren't presented originally when you don't have the original data is a great idea.
Let me elaborate more on what I am doing so you might be able to help me better.
I am studying the implementation of service X, so I created an online survey.. distributed.. get results.. and so come the time to analyse it
The survey website provide me with charts indicating what was respondent's answers but I have to do descriptive statistics for my paper not just presenting these charts or even the descriptive measurements which I mentioned before.. What I am looking for exactly is how can I help the reader understand these descriptive findings? I tried doing many searches and found that running ANOVA test or p value to relate some questions to others might be good option but again I don't know how to do that with the set of data that I have..
If you can explain to me what can I do? or direct me to a helpful video showing how to do the ANOVA assuming independent observations with what I have, I would be really grateful as I already did many searches but all showing how to do the test with raw data.
I managed to get the raw data from the survey website.. but it needs too much work because there are some empty cells in some questions for each respondent so I can't run any test before removing it or make all constants as numbers.
Please Please if any one can just help in analysing by mentioning the mean and variance.. What can I benefit from these measurements as a researcher.
for example: Q2 age:
group 1: 22 respondents
group 2: 42 respondents
group 3: 134 respondents
group 4: 116 respondents
group 5: 26 respondents
the mean = 3.24
standard deviation = 0.99
variance = 0.97
Is right to say that mean shows us that most answers are around group 3? and the variance emphasises that because 0.97 variance indicate that answers are similar and not that much different?
or any suggested tests to be run with the numbers of respondents for each question?
Please do reply ASAP as its really urgent and thanks in advance
"Age" seemingly was categorized (such as "1=18 to 25 years", "2=26-35 years" etc.)?
In such a case (ordered categories), the mean is not useful. You should use the median
instead. In addition, you could make a a chart where you display the relative frequencies
of the age groups (something like this here, or like this, or this).
Thanks a lot for your reply... I already did visual charts displaying the results but my professor asked for more details statistical explanations for the findings.. like relating demographics to other questions
Yes, answers were categorised as you mentioned above
so Is the median in this question = 42? which means the second group? and then say that the answers of question number 2 are centred around group 2? while the majority of respondents belonged to group 3 (134 respondent = 39%) ?
please correct me if I am wrong!
and what about the variance and standard deviation?