Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Help with SPSS

Options
  • 23-04-2007 1:50pm
    #1
    Lucifer
    Moderators, Technology & Internet Moderators Posts: 1,467 Mod ✭✭✭✭


    Hi

    Have never used SPSS in my life until yesterday and starting to get me head around it a little bit. Have set up a few datasheets (not sure what to call them!) and produced a few graphs etc but have now got to a part which i am stuck on. I have to perform Correlate without using the correlate function. I have the formulae(sp?) that i need but am having trouble using compute variable with the formulae. i just cant figure out what way i should be typing them so it will compute the correct data that i want etc. any help would be greatly appreciated. attached is a sample of one of the formulae that i am trying to compute on a column with 473 cases. thanks in advance.

    Edit: Sorry if it doesnt make perfect sense! is all new to me so i'm not the best at explaining it!

    Kev


Comments

  • Options
    #2
    2Scoops
    Registered Users Posts: 1,845 ✭✭✭2Scoops


    It doesn't make much sense tbh:D

    The formula you have is for the standard deviation - why can't you just use the descriptives function? I'm a bit confused as to how you want to populate a new column because there will only be 1 standard deviation for your entire column with 473 cases...


  • Options
    #3
    Lucifer
    Moderators, Technology & Internet Moderators Posts: 1,467 Mod ✭✭✭✭Lucifer


    Calculate the correlation coefficient of Age and Salary without using the statistical procedure 'Correlate'.
    The formulas used can be broken down and calculated using SPSS's compute command into new variables.

    This is what i must do so i must calculate this using compute and formulae to achieve my result. The thing I am having trouble with is how to use some commands in compute to carry out the formulae with the correct data.
    Example:
    MEAN. MEAN(numexpr,numexpr[,..]).
    How do i use this formula to calculate the mean of the 473 cases?

    The formula I posted for standard deviation, is an example of the ones i am trying to compute.

    Thanks again, hope i made a little more sense!

    Kevin


  • Options
    #4
    2Scoops
    Registered Users Posts: 1,845 ✭✭✭2Scoops


    Well, this now sounds suspiciously like getting assistance in a school or college homework so I'm less inclined to help.

    But here's a hint: transpose your columns so that they are in rows instead: the MEAN(numexpr,numexpr[,..]) function will then work and it will just be a matter of building successive formulae. There's probably an even easier way to do it that would be less effort since, technically, you only have to not use the correlate procedure.

    What kind of A-hole-ish assignment is this anyway? I can understand making people familiar with how to calculate simple statistics to gain an appreciation of the maths behind them, but using SPSS in such a clunky way when time would be much better spent learning how to use it properly is stupid and a little sad.:(


  • Options
    #5
    Lucifer
    Moderators, Technology & Internet Moderators Posts: 1,467 Mod ✭✭✭✭Lucifer


    While it is part of a college question, its not my college question, I am trying to learn a bit by completing the question which i have so far until this part of it as I will have to do it in the future so i got a question off someone i know who is in college.

    I understand you dont want to just do the work for me and i wish to solve it myself but am still struggling,

    MEAN(numexpr,numexpr[,..]) what do i enter into this function? numexpr? and what goes where the [,..] is? is it like excel where you can select a range of cells to use?
    also what does this mean? transpose your columns so that they are in rows instead.

    I agree this does seem like an unusual question to ask since there are easier ways of doing this, but i am trying to do it anyway as i feel it may be useful to learn for future? do you agree? is it something that is necessary to know for using SPSS?


  • Options
    #6
    2Scoops
    Registered Users Posts: 1,845 ✭✭✭2Scoops


    Lucifer wrote:
    While it is part of a college question, its not my college question, I am trying to learn a bit by completing the question which i have so far until this part of it as I will have to do it in the future so i got a question off someone i know who is in college.

    Fair enough:)
    MEAN(numexpr,numexpr[,..]) what do i enter into this function? numexpr? and what goes where the [,..] is? is it like excel where you can select a range of cells to use?
    also what does this mean? transpose your columns so that they are in rows instead.

    Transposing the columns means instead of the 473 values going down column, you have 473 values going across in rows. So you will have 473 columns instead of just one. It's a simple matter of [Data drop down ... Transpose...]. Then with the MEAN(numexpr,numexpr[,..]) put in all 473 columns - I know: it's ridiculous!
    I agree this does seem like an unusual question to ask since there are easier ways of doing this, but i am trying to do it anyway as i feel it may be useful to learn for future? do you agree? is it something that is necessary to know for using SPSS?

    I'm 100% confident that this is a complete waste of time - you're spending hours of time computing something that can be done in a few seconds - it's the reason why SPSS was invented. By going through the proper menus, you'll get loads of other useful and relevant information as well. I feel sorry for your friend who is wasting his time with such stupid assignments.

    In addition, only the most simple of correlations will be able to be done in this way - any kind of advanced model and you'll have to go back to proper menus anyway. It's definitely worthwhile calculating one or two simple ones so you get a feeling for the maths behind it, but once you know that, there is no further benefit to doing things the long way.

    Knowing the basic maths, and learning how to use SPSS to it's fullest capacity is a much more worthwhile endeavour - trust me;)

    Actually reminds me of an old professor of mine - he spent his life in the 1960s working out ANOVAs by hand. Spent years doing it with an old calculator and slogging through them one by one. Now his life's work could be done (more accurately!) with the touch of a button:eek:


  • Advertisement
  • Options
    #7
    gormdubhgorm
    Registered Users Posts: 18,260 ✭✭✭✭gormdubhgorm


    This is probably a really stupid question but...

    How do you group and categorize questions from spss into the one variable?

    For example I want to classify two particular questions under "detterance" variable heading

    1 severe fines would deter people from illegally down-loading music?

    2 Improved technology would deter people from illegal music copying and down-loading?

    I want to group the data on these questions under the one variable "detterence"

    How would I do so? I was trying to input some data earlier on today but I realised I was making a variable out of each question individually. Am I doing something wrong? Or can each particular "question variable" be grouped together when the data is completed????

    Guff about stuff, and stuff about guff.



  • Options
    #8
    2Scoops
    Registered Users Posts: 1,845 ✭✭✭2Scoops


    gormdubhgorm wrote:
    This is probably a really stupid question but...

    How do you group and categorize questions from spss into the one variable?

    For example I want to classify two particular questions under "detterance" variable heading

    1 severe fines would deter people from illegally down-loading music?

    2 Improved technology would deter people from illegal music copying and down-loading?

    I want to group the data on these questions under the one variable "detterence"

    How would I do so? I was trying to input some data earlier on today but I realised I was making a variable out of each question individually. Am I doing something wrong? Or can each particular "question variable" be grouped together when the data is completed????

    It depends on how you want to analyze the data. I'm not sure what the most correct way to do it is, in this case. If you're already sure that you can group your questions into a latent variable, or if it's just a relatively simple school project, you can try:
    1. write it in 'long' form - so the same person's responses end up in different rows, with grouping variables to label the questions and which ones are linked, then use linear mixed models to analyze
    2. write it in normal form (each question gets its own column) and use some kind of repeated measures design [not sure if that's allowed with this kind of questionnaire stuff though:)]
    3. Use AMOS and then you'd be able to get everything the way you want it but you may not have it installed (it's an SPSS add-on) and it's a little difficult to use without instruction.

      OR
    4. Simply make a new variable called 'deterrence' that has a score based on the responses to the 2 questions [the easy way out but effective nonetheless:D]

    If you're actually trying to develop a reliable and valid questionnaire, then go with exploratory factor analysis.


Advertisement