• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

Would you help me fill out a survey? (For my education!)

patrickpetersen

patrickpetersen

Newbie
Hello guys and girls :)

I'm at my last year at the University of Southern Denmark in Esbjerg (DK). And for an exam I wrote a month ago, I've developed my survey a little, so I can make a better analysis.

The survey is about smartphone and usage of the smartphone. At the last part (The boring part!!) I've made something called an orthogonal design. (more info in the survey)

I would be happy if some of you wanted to use apporoximately 10 minutes to make my education better. :)

Thanks in advance!


Link to survey : https://www.survey-xact.dk/LinkCollector?key=TFTPAH99CK3P

- If you see some typos, please PM me :)
 
B2L said:
I just finished it as well, my only question is; why the odd smartphone configurations?
Click to expand...

Thanks!
Well it's called an orthogonal test design. I define some values an attributes and a program called SPSS mixes it all together in total randomness. When I saw them I was like.. "wtf?!".

But the individual design doesn't matter. It's because it's for a conjoint analysis to find which kind of things that are more important than others.

An example could be:
You want to buy the new SGS3. When you go down to the store you have two options:

- SGS3 16 GB.
- SGS3 32 GB.

You can get the 16 gigs now.. or you can wait one week for the 32 gig version.

If you choose the 16 gig version, it indicates, that storage space isn't as important as you having that phone right now.
The other way around, if you choose the 32 gig and waits a week, it indicates that it's important for you to have a lot of storage, but less important to get the phone right now.

That's with two variables! You can make it with 100 variables.. I choose 5 main variables.. it gave me 25 test designs.. that should be enough to make a clear image.. :)

I hope you understood me .. if not.. just ask again :)
 
patrickpetersen said:
Thanks!
Well it's called an orthogonal test design. I define some values an attributes and a program called SPSS mixes it all together in total randomness. When I saw them I was like.. "wtf?!".

But the individual design doesn't matter. It's because it's for a conjoint analysis to find which kind of things that are more important than others.

An example could be:
You want to buy the new SGS3. When you go down to the store you have two options:

- SGS3 16 GB.
- SGS3 32 GB.

You can get the 16 gigs now.. or you can wait one week for the 32 gig version.

If you choose the 16 gig version, it indicates, that storage space isn't as important as you having that phone right now.
The other way around, if you choose the 32 gig and waits a week, it indicates that it's important for you to have a lot of storage, but less important to get the phone right now.

That's with two variables! You can make it with 100 variables.. I choose 5 main variables.. it gave me 25 test designs.. that should be enough to make a clear image.. :)

I hope you understood me .. if not.. just ask again :)
Click to expand...

That completely makes sense now, I was just interested. I saw one question that said would you buy a phone with iOS, 4.8" screen, and a 1 GHZ single core processor; it made me think wth?
 
B2L said:
That completely makes sense now, I was just interested. I saw one question that said would you buy a phone with iOS, 4.8" screen, and a 1 GHZ single core processor; it made me think wth?
Click to expand...

Yeah, It had some wild designs.. I think there is a 5.3" android with quadcore with a price less than $150 :p If that was real.. I would buy 3.. :p

I hope to get around 500-1000 respondents to get the big picture :D
 
Wilkas said:
Done, glad to help... Any chance you will be sharing the results on here? I'd be interested in the results of some of those questions.
Click to expand...

Thanks! Yeah , it could happen. I have to write it in Danish first.. and then translate it :) Is there a specific part your especially thrilled about? :)

I just need enough respondents , right now I have like.. 10 % done :)
 
gbiggie said:
Would like to help you out but there is no option for my primary device:(
LG
Click to expand...

I'm thinking of a solution! :) I've received some notes on that one. The problem is, that for 50+ answers there will be a hole.. but I'm thinking of doing something like filling out the possible blanks with the mean value.. I just have to find out how in the system.. That's the most used method, but it's hidden well :p
 
I did it, but my device configuration answers tended towards extremes: either highly likely or no chance, not much inbetween, with veto usually based on whether some required spec was failed (too little storage or unacceptable OS being main ones)
 
Hadron said:
I did it, but my device configuration answers tended towards extremes: either highly likely or no chance, not much inbetween, with veto usually based on whether some required spec was failed (too little storage or unacceptable OS being main ones)
Click to expand...

It's okay to have extreme cases.. after all, it's still a matter of a conjoint analysis, so it's nit a problem at all :-)
 
Android and Windows Phone options, I look at the specs. Blackberry, Symbian and iOS, instant Will Not Buy, lol! I'd love to see the results. Have you posted this anywhere else though, like Apple or Blackberry forums? You're going to get a very biased result if only posted here!
 
Hrethgir said:
Android and Windows Phone options, I look at the specs. Blackberry, Symbian and iOS, instant Will Not Buy, lol! I'd love to see the results. Have you posted this anywhere else though, like Apple or Blackberry forums? You're going to get a very biased result if only posted here!
Click to expand...

I have to post it as those forums as well, you're completely correct! Do you have some forums which are good posting at? I've only some danish ones with low activity! Don't know some good ones (I'm a Droid too ;) )

But the Orthogonal design with the specs thingy, isn't only for the conjoint analysis. It'll show which specs is the most important ones for the respondts :)

Usually they do not make sense, but if you fx want to buy android with small screen but high cpu power, and the next is android with low cpu and small screen.. it shows the the OS android is an important factor of choise :-)
 
You must log in or register to reply here.
Back
Top Bottom