Help for the SPSS greenhorn

[knowing]

Thank you in advance for any and all help with these questions.

How do I form a new data set directly in SPSS with the info(question and corresponding answers sets from a survey formulated in survey Monkey?

Can I have available choices different for each question within the same survey? For example:

1. Are there any other learning factors that affect your child's performance academically?(would this be Chi Square or can I just envelope in this to my data set as is?)

a. Yes =1
b. No =2

2. How would you rate your child's performance since diagnosis?

a. Excellent =1
b. Above Average = 2
c. Average =3
d. Below average=4
e. Very poor=5

Is the coding on the right track?

---

Members don't see this ad.

 

[cmuhooligan]

Thank you in advance for any and all help with these questions.

How do I form a new data set directly in SPSS with the info(question and corresponding answers sets from a survey formulated in survey Monkey?

Can I have available choices different for each question within the same survey? For example:

1. Are there any other learning factors that affect your child's performance academically?(would this be Chi Square or can I just envelope in this to my data set as is?)

a. Yes =1
b. No =2

2. How would you rate your child's performance since diagnosis?

a. Excellent =1
b. Above Average = 2
c. Average =3
d. Below average=4
e. Very poor=5

Is the coding on the right track?

You can import data from surveymonkey into SPSS; however, I believe you need to first open the data in excel, then you can import the excel file into SPSS.

Yes, you can format your survey so that each item has different response choices.

Whether you would use a Chi-square or not is a question of data analysis, not survey design, but to answer your question, the way the item is phrased could be easily imported into SPSS.

 

[Ollie123]

In order to get data from survey monkey to SPSS, you should export off the website into Excel. I believe there is some option (can't remember what) where you tell it to give you numbers instead of text. Otherwise it will output the text response for every variable and you don't want that.

With few exceptions, the numbers will range from 1 - ? depending on how many answers there are - it goes in the order they are displayed so you will need to recode them if you want it done differently.

Truth be told there is no "Right" way to code. You could make yes -667 and no 7,901,436 and it shouldn't make a difference if its being analyzed at the item level. I always do No = 0, Yes = 1, but that's just convention - when hand-entering measures it helps to have a system so everyone knows to enter the same thing for "No" every time. I'm convinced people flipping around their coding on Yes/No is responsible for more than a few spurious findings in the literature.

I'm not sure what you mean by "Envelope this to your data set". If you are interested in the item itself, yes you shouild probably be using chi-square. Well, not necessarily depending on the question of interest, but the point is it needs to be an analysis that treats it as a categorical variable.

Item 2 - coding is fine, but again, it depends on how you plan to analyze it. Is the item being looked at on its own? In that case, again, you shouldn't run an ANOVA since it isn't really a continuous variable. If its being scored (i.e. summed, averaged, etc. for a scale) then you can justify treating it as continous and run an ANOVA or solmething like that.

 

[knowing]

You can import data from surveymonkey into SPSS; however, I believe you need to first open the data in excel, then you can import the excel file into SPSS.

Yes, you can format your survey so that each item has different response choices.

Whether you would use a Chi-square or not is a question of data analysis, not survey design, but to answer your question, the way the item is phrased could be easily imported into SPSS.

Thank you so much..We will be running analysis in depth on some of the variables later. For the moment we are just concentrating on the importation of the data into SPSS. The chi square question I am just importing with the rest for now.

 

[knowing]

In order to get data from survey monkey to SPSS, you should export off the website into Excel. I believe there is some option (can't remember what) where you tell it to give you numbers instead of text. Otherwise it will output the text response for every variable and you don't want that.
I will look for the command tonight in SM for the exportation

With few exceptions, the numbers will range from 1 - ? depending on how many answers there are - it goes in the order they are displayed so you will need to recode them if you want it done differently.
This is good because I was worried that a question coded 1-5 and a question coded 1-15 within the same survey would give a crazy result.

Truth be told there is no "Right" way to code. You could make yes -667 and no 7,901,436 and it shouldn't make a difference if its being analyzed at the item level. I always do No = 0, Yes = 1, but that's just convention - when hand-entering measures it helps to have a system so everyone knows to enter the same thing for "No" every time. I'm convinced people flipping around their coding on Yes/No is responsible for more than a few spurious findings in the literature.
I'm coding the same as you 0 1 2 3 and so on so were good there.

I'm not sure what you mean by "Envelope this to your data set". If you are interested in the item itself, yes you shouild probably be using chi-square. Well, not necessarily depending on the question of interest, but the point is it needs to be an analysis that treats it as a categorical variable.
No, not the item itself at the moment that will come later. I can treat this the same as the others for now because it is a one group sample descriptive...so were good there.

Item 2 - coding is fine, but again, it depends on how you plan to analyze it. Is the item being looked at on its own? In that case, again, you shouldn't run an ANOVA since it isn't really a continuous variable. If its being scored (i.e. summed, averaged, etc. for a scale) then you can justify treating it as continous and run an ANOVA or solmething like that.

It is only one group descriptive so no for the ANOVA. I think my prof wants to change the depth of the analysis focused on each question later depending on the target audience and whom we will be writing the article for.
I so very much appreciate your help..your a lifesaver...

 

[knowing]

I don't know why it put some of my responses to you in the blue quote part and the other response at the very bottom???

 

[leavingprov]

Just want to let everyone know about a fanTAStic book I've been using at work for SPSS, called "SPSS Survival Manual" by Julie Pallant. It's super well-written and step-by-step. I have been borrowing it from a co-worker but I'm definitely planning on picking up a copy during grad school!

Good luck budding statisticians!

 

[knowing]

Just want to let everyone know about a fanTAStic book I've been using at work for SPSS, called "SPSS Survival Manual" by Julie Pallant. It's super well-written and step-by-step. I have been borrowing it from a co-worker but I'm definitely planning on picking up a copy during grad school!

Good luck budding statisticians!

This is so sad(laughing snort)...my husband ask me what I wanted for my birthday tomorrow and I told him to order me this exact book from Amazon. How psych nerdy is that?

 

[leavingprov]

Your post calls to mind the photo in today's NY Times article :

http://www.nytimes.com/2009/08/06/technology/06stats.html

 

[knowing]

Your post calls to mind the photo in today's NY Times article :

http://www.nytimes.com/2009/08/06/technology/06stats.html

I'm loving that article...I printed it and it's posted on the frig...let's say for encouragement purposes...Thanks!