Do you like this blog?
Isn't that a easier question to answer than a question like "How would you rate my blog on a scale of 0-10?"
Am I not bound to get more truthful and accurate answer with the first question? After all, our minds can easily distinguish whether we like things or not, but it is very difficult to quantify your liking/disliking for something.
If I have 100 people taking a survey and then rating it 0/1, I can still come up with a 0-10 figure in the end by just taking a mean of the answers and multiplying by 10. Would that figure be any less or any more accurate than an average derived from making people give a 0-10 answer?
Feeding a 0-10 answer obtained by averaging 0/1 answers into a larger calculation might still work, but I think the basic inputs should always be binary.
Subscribe to:
Post Comments (Atom)
2 comments:
Oh man, can't believe you are touching upon the category vs continuum debate!
Frankly, it all depends on the type of data (feedback from readers in this case) you want to exploit (statistically speaking)
If you simply pose the question - Do you like this blog, you are making a hell of a lot assumptions that the reader understands the meaning of "like" in this context (which can be numerous)
You can ofcourse pose numerous questions such as:
Do you like this blog in terms of Content?
Do you lile this blog in terms of design?
Do you like this blog in terms of frequency of posts?
This is helpful if you want an "impulse" answer from the reader.
(Most readers when presented with surveys just want to get done with the questions)
In the continuum side of things one can really understand what the data actully means. Let's take the same example of Do you like this blog and lets set the likeness value -
1 being I absolutely hate the design and content of this blog and can't wait to get out of here.
2 - The design is terrible, strains the eye and the content is okay but I like to be updated on what is Akash upto
3 - ...
.
.
.
10 - I love it, the design is the best I have seen and the content is great...
You see my point, in the continuum side of things you kind of control the context of "like" and IMHO makes the reader think about his/her answer choice.
It really depends on a lot of factors and there are trade-offs to both approaches.
Just my 2 cents!
My point is that even if you want the reader to think of the answer, the reader is not capable of making that decision in 99% of cases unless he/she has an expertise in that area. Whereas, if it is just like/dislike, he is always going to make the correct answer.
I remember this point being made in some book somewhere. I will search and try to find out.
Post a Comment