Docstoc

SPSS

Document Sample
SPSS Powered By Docstoc
					                                      SPSS
                                          Part 1:
Data Entry and some VERY simple descriptive stats
    SOME WORDS OF WARNING…
 A reminder that you will be using SPSS next semester in Psyc
  310, in your own research, if you serve as research assistants,
  and possibly in other classes in the sciences (Bio).
 We will NOT be re-teaching you the use of the program each time
  – you are expected to pay attention the first time.
 There will be 3-4 homework assignments associated with SPSS
  this semester
   … The first two will count as independent homeworks (data
     entry and analysis) and there will be an in-lab “test” on a
     common dataset during our last scheduled lab period which
     will contribute 5% to your total class grade. Some future
     ‘generic’ homeworks will also include SPSS tasks, but the
     ones identified on the coursepage in red are the only ‘pure-
     SPSS’ assignments.
    GETTING STARTED
   Open SPSS


   Likely be choosing:
    …   “type in data” (New Dataset)

    …   “Open an existing data source”



   Since we are creating a dataset,
    select the “type in data” button
    and hit “OK” at the bottom.


   Once that window disappears
    you will see that you are in
    either the “data view” or
    “variable view” page
    DATA VIEW PAGE
   This is where you will see
    the data in its’ raw form, and
    the page from which you will
    be running your analyses.



   The page that you are on will
    be highlighted in yellow
    down in the bottom left of
    your screen.



   We need to start by defining
    our variables for the
    program … and for that we
    need to be in “variable view”.
    …   Click on the tab for variable
        view to switch over
      VARIABLE VIEW PAGE
   As you can see – we are going to define our characteristics by what
    Kind of data we have (NOIR) and put in labels so that we can
    understand the printouts.




Each of these
values represents
one data column
on the data view
page.


It is easiest if you
know all the data
you will have to
enter beforehand
so when you set
this up, it is
organized
                                 Quiz Average for
                                  student (0-100)
OUR DATA
                                 Exam Average for
                                  student (0-100)


  We will pull a chunk of       A data grouping
   our data from my               answering this
   classpage …. And some          question: “What was
   of it you will be making       the higher percentage –
   up.                            quiz or exam?” (Quiz or
                                  Exam)

  First though  What
   kind of data do we            Your ranking of which
   have?                          name you like best
                                  (rank from 1-10)


                                 GPA (0-4.00)
                                    Quiz Average for student
WHAT KIND OF                         (0-100)
                                     …   Ratio
DATA?
                                    Exam Average for
                                     student (0-100)
                                     …   Ratio
    What kind of questions
     would we be able to ask/       A data grouping
     answer with the data            answering this question:
     that we have?                   “What was the higher
                                     percentage – quiz or
                                     exam?” (Quiz or Exam)
    We can organize the data        …   Nominal
     any way that we’d like to
     (e.g. by the most              Your ranking of which
     important to least              name you like best (rank
     important) as long as it        from 1-10)
     makes sense.
                                     …   Ordinal

                                    GPA (0-4.00)
    We will organize by type        …   Ratio
     of data (NOIR)
                    NOMINAL DATA
Start by entering a name that ‘captures’ what our nominal variable stands for
                   “Highscore” --- Note only use one word




       Remember that in our experiments – these will likely be the IV’s



 Notice that when you hit enter – a number of other columns get filled in.
 These may not be correct for what we need (the computer cannot read
 your mind) – if you don’t enter the data in correctly, mistakes WILL be
 made in the analyses. So next, click on the “MEASURE” icon
                      Choose nominal



      Next, go to the “Values” column and click on it and a new screen pops up




Add in a Code for Value (must                                          Finish adding all
be numeric). And a label you                                        values/ labels then click
will understand                        Click on “ADD”                          OK
                     ORDINAL DATA




As with the Nominal data:
        First Enter in a Name for the variable in column 1: RANKING

        Then go to the “Measure column” and select ORDINAL

        DONE!
                There isn’t any other step as there are no groups to label
               INTERVAL/RATIO DATA




Our last 3 pieces of data are all Interval/Ratio (remember that mathematically,
these are going to be treated the same and do not need special separations for
SPSS.

Set up the last 3 – QUIZ, EXAM, and GPA.
         This time, however, we do not have to make ANY adjustments in the
“Measure” column. So just hit enter, then move to the next.



Now we’re ready to enter the data, go down to the
bottom left and click on the “data view” button
    The 5 lines of data we entered are now our column headings in the data view
    page. Now we enter our data….




            ONE THING THOUGH – we will enter in the value
            labels (1,2) in column 1 rather than the words Quiz,
            Exam




All data except for the “Ranking” and GPA comes from the webpage I sent.

                            Enter the data.
                                OOPS!

                                  This actually is an accident… but
                                  this would fall under the category
                                  of “Hmm, I should have thought
                                  about that when I was designing
                                  the project. I now need a rule to
                                  deal with this.

                                  However, SPSS (if left alone) will
                                  run the analysis excluding this
                                  piece of data whenever Col 1 info is
                                  used




Now enter in your scores for the ranking and GPA – they need to fall into the
appropriate parameters (a GPA cannot be 7.23). But make up the data.
Here’s the data I have (yours will
look a little different in col 2 and 5).

Let’s say we want to clean up the
data a little bit
         In the ranking column, it
doesn’t make sense to have decimal
points (nor in the first column)

** We’ll head back to the Variable
view by clicking on the tab at the
bottom, and changing some of those
boxes we didn’t use the first time.
Decimals = Change the values here to change what shows up on the data page
       Change Highscore and Ranking to “0” decimals

Columns = the number of characters in the column headings

Align = just like in word to make things more readable to you
 Now – why don’t we save what we have.

Use either the icon or pull-down menu “file” and save
This is just an annoyance that shows up after each action – close or ignore if
it isn’t an analysis.
  NOW LET’S MAKE SOME CHANGES
Imagine that, to our surprise, we forgot an entire column of data (or in this case we
had data that only became available later).

How do we change what we have?

Let’s say our new piece of data is an overall class ranking (their ranking in the entire
class or 20XX) not their ranking in this class – and that data is helpful to us.

         (1) Because it is ranked data – we want to keep it next to the other ordinal
data so we have to insert a column.

          (2) Because we used the label ‘ ranking ’ when we entered the data – it is
more likely to confuse the two columns we have so we want to change the column
names. Remember that these two are both rankings – but are ranking VERY different
things.

         (3) Because the rankings are from the entire class (of let’s say 200 people) the
rankings seem incomplete and we need to re-rank the scores.

          (4) One student dropped the class (or we were forced to drop her because of
the tied percentage) – so we need to replace her scores with another student
  INSERTING A COLUMN
First off – switch back to “variable view.” Go to the pull-down menu under “EDIT”
and select the “insert variable” option.
 As usually happens when you use an “insert function,” the column may have been placed in
 the wrong place (you want it after the original “ranking”)




To delete a row – click on the very first column on that line (the numbers) and the row will be
selected. When it is highlighted – hit the delete key.




Place the cursor on quiz and insert variable again – so SPSS automatically inserts ABOVE
The name is highlighted so we need to give it a name and change the scale to
                                 Ordinal




                 1) Change the name to “20XX_rank”
               2) Change the final column to “Ordinal”
          3) Rename variable 2 from ‘Ranking’ to “NameRank”
           4) Remove the decimals option for the new column
Here is the new data. Notice that the other columns moved in response to
 our changes on the other page. We now need to remove the one piece of
            faulty data – the one with the tied exam/quiz score.

Select the row of data by clicking on the number all the way to the left and
                                 hit delete.

 Now enter in the data for Uzume (we will assume that she has the same
                      “name ranking” as did Freya
Here are the overall class rankings



            With a small dataset like this, we
            can easily do the rankings yourself.

            With a large dataset (and
            numerable, potential tied rankings
            etc.), it might be better to let the
            computer do this for us.

                      Select the column by
            clicking on the column heading
            (turns blue), go up under the
            “TRANSFORM” pull-down menu –
            select “rank cases” and you’ll get
            what shows up on the next page…
(1) Select column to be ranked



                                           (2) Shift it over to the
                                           “variable(s)” box by clicking
                                           on the arrow




                                            (3) Double check the order in
                                            which the ranking is to be
                                            done – Assign rank 1 to
                                            lowest value.




                      (4) Finally, click on “Ok” button
New column has
been created for us
                                   I cleaned mine up by deleting
                                   the original data column
                                   (remember that this an act of
                                   finality – once lost the data is
                                   gone).

                                   I then moved the newly ranked
                                   column over (and took out the
                                   decimals).




NOW …. Let’s do some statistics!
    DESCRIPTIVE STATISTICS
   We are going to
    collect:
    …    all the various
        descriptive
        statistics we’ve
        learned about
        so far

    …   a set of
        frequency
        distributions
        for our
        variables
(1) Select all 3 of our Interval/Ratio variable and shift them into Variable(s)
column




                                                                   (2) Select the
                                                                   Statistics button
                                                                   to bring up all of
                                                                   the available
                                                                   options.
                                                                   By doing this…
                                                                   we do not have
                                                                   to run the
                                                                   “descriptives”
                                                                   analysis.




                    (3) Select the options shown and then hit
                    “continue” and you’ll be taken to an output
                    window with the statistics done
Statistics
                                    Quiz               Exam      GPA
N            Valid                  10                 10        10
             Missing                0                  0         0
             Mean                   75.1300            74.6700   2.8000
             Median                 74.1500            74.5000   2.8400
             Mode                   58.80a             75.90     1.73a
             Std. Deviation         9.72603            9.00704   .67082
             Skewness               -.046              .078      -.134
             Std. Error of Skew     .687               .687      .687
             Kurtosis               -.482              -.592     -.928
             Std. Error of Kurtosis 1.334              1.334     1.334
             Range                  30.40              29.10     2.08
             Sum                    751.30             746.70    28.00

a. Multiple modes exist. The smallest value is shown




       You CAN of course – print things up …. BUT PLEASE REMEMBER …
       you could also just keep the info on your computer and save a few trees.

       Print only what you need to turn in.
Quiz
                         Frequency Percent   Valid Percent   Cumulative Percent

Valid   58.8     1       10.0                10.0                10.0
        64.2     1       10.0                10.0                20.0
        70.8     1       10.0                10.0                30.0
        71.7     1       10.0                10.0                40.0
        73.3     1       10.0                10.0                50.0
        75       1       10.0                10.0                60.0
        77.5     1       10.0                10.0                70.0
        82.5     1       10.0                10.0                80.0
        88.3     1       10.0                10.0                90.0
        89.2     1       10.0                10.0                100.0

        Total    10      100.0                                   100.0




   Good news! You’re now ready to do the first of the computer
   homeworks.

   Check the webpage for due date and requirement.

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:6
posted:8/11/2012
language:
pages:30