Thank you for taking the time to comment. I can't find where I state that income/volunteer for organization is interval. But you are exactly right. If they are unequal groups, with a hierarchy, then it would be an ordinal variable.

Qualitative Questions are open ended questions that allow the respondent to freely write out their answers. SPSS will classify them as a String variable, which means that you cannot run any real mathematical analysis on them. However, you have several options for analyzing these questions - most of which do NOT use SPSS. To use SPSSS, you will start by reading through all the responses and identify general themes that multiple people mention.

Create a dummy variable in SPSS for each theme that you identify. Then code your data by giving each case a "1" if they exhibit the theme or a "0" if they don't. Then you can run frequency analysis on each variable you create. If you are not sure how to run the frequency analysis and how to create charts - then watch my Multiple Response Nominal variable video to see how b/c the dummy variables are essentially a multiple response nominal variable that you have created. I hope I answered your question. If not, please let me know and I will gladly help!

It is ratio. It is an open ended question, with numbers, and it has a zero as an option. An interval question is one where the researcher creates a scale that has equal intervals. Interval/ordinal/nominal questions always have scales that are created by the research. Ratio variables are typically open ended.

@@JenniferTaylorPhD Thank you for your response. Really appreciate it!

I would contact Qualtrics support. I have had an instance where there was a coding glitch and tjey had to clean up a coding error that was caused by its auto save function.

Hi did you find any resources on this?

@@waynemwangi9444 unfortunately no

Hi Jenny! Just to summarize, you want to use the net promotor score (Qualtrics will calculate it for you) as your dependent variable (what you are predicting) and you have constructs that are measured with multiple measurement items. In order to run a regression in SPSS you will need to create index variables for each of your constructs. There are two primary methods you can use. You can either create a variable by computing the sums of the answers from the measurement items, or you can create a variable that computes the average score of the measurement items. Which one your choose depends on the variables you are measuring. You will create these variables in SPSS, not Qualtrics. This video will show you how to compute a new variable: kzbin.info/www/bejne/fH6keoqsa6yae8U

@@JenniferTaylorPhD Thank you so much, got it! One more question: I have 2 dependent variables and 4 independent variables, and want to test 1 of the dependent against 2 of the independent (and the other one against the other two)- what type of regression is the best? multiple and run it twise? or multivariate?

And one last thing - when I have summarized the variables into one variable to use it in a regression analysis, do I need to show how i did this/what the average is when I present the findings, (in a dissertation) or is this not necessary?

@@jennykristinebekken1790 It depends on the variables - are the dependent variables continuous or categorical? If they are categorical, how many categories are there?

@@jennykristinebekken1790 You will describe your method for creating the new variable in the measurement section of your Methods section. For example, if you had three variables that described "Service Quality" then you would have a subsection in the Measurement section that described the measure, the questions you used to measure the construct, and then a description of how you created a composite (average) variable that you used in the analysis.

Hi Israel! Thanks for your question. If you asked one open ended question and the answers could fall under multiple service categories, then I would do the following: 1) Create a column for each of the service categories that are mentioned by resondents (ie., food court, payments, enrolling, parking, mobile app). Essentially these service categories become "themes" - but they are "higher order themes" in that you will create subthemes underneath them. Due to the nature of the question, you might find that a respondent provides feedback for multiple service categories (maybe they talk about the food courst, parking and the mobile app). Simply code a 1 in each corresponding column for the service category. 2) If you are using excel, you can create a new worksheet for each service category and copy and paste the respondents answers that you coded as mentioning this category. Then, assuming you have enough responses, go through an identify any themes that you see in the data (for example maybe those responses that mentioned food court talk about things like slow service, cleanliness, etc.). Themes are a way to summarize data, so if you don't have enough responses to identify themes, then in your report you can just report the distinct issues that were mentioned and you can report those verbatim. I hope this helps! Good luck with your project :)

Hi Karis! Thank you for your question. Without understanding the context that your professor is referring to, I cannot really offer an opinion. This issue is that there are a multitude of methods for qualitative data analysis and which method is best depends on the research question and the data collection methods (open ended survey questions, interviews, focus groups, etc.). In the video I present a very basic thematic analysis that is used on open ended questions from online surveys. A more rigorous thematic analysis is not necessary in this instance, as the research is descriptive and we are not trying to develop theory. Additionally, this analysis is not for qualitative data that has been collected as the sole means of analysis. In this case, the basic thematic analysis is simply adding to the story that the quantitative data tells us. The type of data analysis I present may be used to identify future measurement items (survey questions and categorical answers). It is important to note that thematic analysis, by its nature, is about binary coding - either the data exhibits the theme or it does not. When you have a survey question and you identify several themes of answers, you can use graphics to show these themes. However, it can be argued that it is subjective and thus the percentages obtained will reflect bias and error, which is why many do not recommend using graphics. My perspective is that the graphics add value when conducting descriptive research - such as surveys/questionnaires that include open ended questions. I would not recommend using this method to those who are using qualitative data as a sole means of research whose aim is to be published in an academic journal. Hopefully this helps :) Best of luck to you!

It really depends on your measurement scale. I think this video might help you determine the type of measurement scale you are using for your question type: kzbin.info/www/bejne/faO5l6ODo62pd5Y