Spreadsheets are commonly used for questionnaire analysis, presentation of statistical information, statistical tests, accounts etc. etc.  because they are designed to:

i) manipulate numbers, creates percentages etc.: .
ii) present data in tabloid form.
iii) present data in graphical or chart (Pie and Bar) form.
iv) sort rows of data on basis of number, alphabetical order etc.

II.    Advantages of Spreadsheets like EXCEL & Microsoft WORKS over SPSS.

The advantages of general spreadsheets over statistical packages like SPSS for numerical analysis is that:

i)      Since they are commonly already available, no separate software has to be bought.
ii)     The spreadsheet is easier to enter and reorganise data
ii)      The user can personalise the spreadsheet file (hereafter called worksheet) to give the precise information
         required and make the process more user-friendly.
iii)     The spreadsheet data can be used to create tables/graphs/charts and pie diagrams which can themselves
        be directly copied into wordprocessor files.
iv)    It is easier to understand what is happening. SPSS is very much a ‘black box’!

The disadvantages of spreadsheets over statistical packages like SPSS for numerical analysis is that:

i)      Tests like MANOVA, MANCOVA and descriptive analyses like factor and cluster analysis
        cannot be very difficult to carry out
ii)     Although EXCEL can carry out a range of statistical tests, worksheets are available for a series of tests,
        see below.  The size of the worksheet may have to be adapted to cope with large volumes of data.
iii)     SPSS can read and analyse a complex numerical database very quickly.  It is important to
        recognise that all common spreadsheets can easily create spreadsheets that SPSS can analyse.

 

III.    Tests available on Worksheets available from G.E.Tall*

1. Interval data (eg marks/grades)

Worksheet name Test

ANVAR.***         One-Way Analysis of Variance (t test). Compares the means of 2 or more groups of numbers.
                                                Will use correlated and uncorelated tests.

                                F test Compares the standard deviation of 2 groups of numbers

ANV2WAY.*** Two-way Analysis of Variance. Compares mean scores of both 2 or more groups of students
                                                which are also classified (say) by gender. Assumes that the number of individuals
                                                 in each sub-group is comparable.

REGANAL.*** Regression Analysis. Calculates the correlation coefficient for two groups of students using
                                                initial  and final scores and uses this to plot the regression line (sometimes called
                                                the prediction line or line of best fit). If the line for the experimental group is
                                                outside the confidence interval of the control group then the difference is
                                                 significant @ 5%. Graph of all lines are drawn. Checks homogeneity of lines
                                                (that line fits the scattergram)

2. Nominal Data (Category or headcount numbers)

Worksheet name Test

PROBAB11.*** Sign test Maximum = 30

1WAYCHI6.*** One-way Chi square for 6 categories. The test also reduces the 6 categories to 3 & 2 and
                                            reports the level of significance for each test.

2WAYCHI6.*** Two-way Chi square for 6 categories. . The test also reduces the 6 categories to 3 & 2 and
                                            reports the level of significance for each test.

3. Reading database files:

i)      the number of 1’s, 2’s, 3’s etc.
ii)     the means, median, mode, maximum, minimum and standard deviation of each column
iii)     analyses the data as 1WAYCHI6.*** above and reports the percentage of each category and level of
        significance.
iv)    analyses the data as 2WAYCHI6.*** above

DATAANAL.*** This reads a single column (field) in a database and analyses it. Stating:

CORRMAT.*** Carries out the product moment correlation of ten columns of numbers against the column in
        column Y. This is the first stage of factor analysis.

* (c) G.E.Tall. These files are free whilst studying for your degree at this university.