From the outcome of my survey, I am required to group the respondents based on the Diffusion of Innovation theory, which will place my respondents into different categories.
I am not sure how to achieve this. Currently I am at the early stage of conducting my research.
Would appreciate if anyone can shed light on this.
Thanks
In the Diffusion of Innovation Theory, the main variable for clustering of participants is the time of adoption of innovation.
To do this in your survey, you should have asked participants about the time of adoption. Those who adopted the innovation first (in 2,5% on the left side of the Gaussian distribution), you would put into a cluster of innovators, then next 13,5% of participants you would put into a cluster of early adopters etc. (check the image: https://upload.wikimedia.org/wikipedia/commons/1/11/Diffusion_of_ideas.svg)
If you haven't asked your participants about the time of adoption of the innovation, you should do that to make the clustering.
Dear Muhammad, your research is going to be interesting. If you are going to Adopt Rogers Diffusion theory, I suggest you design your survey instrument which will categorise your respondents into the various segments as proposed by Rogers e.g early adopters, early majority,late Majority and the laggards.Secondly, you need to look into literature to find various studies to identify the variables that were used to measure each segment. This means your survey instrument will have several segments which may be application to each respondent. I hope this helps. Good Luck
Dear Brian,
Yes, it'd be a cluster analysis. Based on the DOI theory, there are 5 categories (innovators,early adopters, early majorities, etc.). Yet, I am not sure on what basis or main characteristic to cluster them accordingly into the mentioned categories. I am still browsing through literature.
Dear Tomislav,
Thanks for the spot-on answer. If time of adoption is the only variable that determines which cluster the respondents belong to, then it should be okay. I have yet to include it in my survey.
Dear Victor,
Yes, you are absolutely correct. I am still looking into the right variable to segment my respondents based on the DOI categories.
Thanks everyone for the help, really appreciate it.
each of your five categories is probably a latent variable. So you approach the problem in that manner
Kind regards
SN
Diffusion theory says people "self select" along a dimension of risk, which means for any given innovation, a person will evaluate the level of perceived risk and then decide how early or late to adopt that new innovation. So the "characteristic" you are looking for is RISK. And of course, the Central Limit Theorem creates the shape of the bell curve and the percentages in each category. I'm not sure that clustering applies here.
- Badges
- Science topic
- Similar topics
- Statistics
- Survey