I have created a simple form to calculate the individual internet continuity needs (will be in percentage) to cluster the customer on the basis of factors (internet continuity needs and amount spending capacity) using K means Clustering. The scale are ordinal form. Since one cannot sum up ordinal scaled data how can I convert the internet needs into the percentage for each individual. Or suggest some ways to design my form in the interval or ratio scale so that i can sum up data values. Link for the form.
https://docs.google.com/forms/d/e/1FAIpQLSeh2NO-QgCaKLkwuPVsnegNNmjKyxLUaQg9pMOgAe7YdJWfCw/viewform?usp=sf_link
Rabin - I can't quite understand your intention here. Your collected sample data is designed to express summary distribution (central tendency and/or variability) across the whole data set. Why do you want to bring the data back to the individual element/data point?
Dean Whitehead I am trying to clustering the customers according to factors(individual internet users needs and amount spending ) using k means clustering. so for this my data points is (x, y) where x is individual internet users needs(on percentage scale) and y is amount spending in Nepali currency(NRS). here i want x to be exact number. since i designed the form in the ordinal scale i am not sure if i can sum up the values.
I don't know what is in your mind or what exactly you want to do but to convert the nominal scale to other scales or vice-versa is rather easy. For example, A has IQ of 100 and B has IQ of 125,
Nominal Scale: A is lower than B,
Ordinal Scale: B is highest and A is lowest,
Interval Scale: The IQ of A is 25 % lower than B,
Ratio Scale: 4 : 5
- Badges
- Science topic
- Similar topics
- Computer Science
- Computer Communications (Networks)
- Internet