Data Analysis with Microsoft Excel: Updated for Office 2007

(Tuis.) #1

310 Fundamentals of Statistics


Use Excel’s CHITEST function to deter-
mine whether this is enough evidence to
reject a hypothesis that the die is true.



  1. Why should you not use the Pearson
    chi-square test statistic with ordinal
    data? What statistics should you use
    instead?

  2. The Junior College workbook (discussed
    earlier in Chapter 3) contains informa-
    tion on hiring practices at a junior col-
    lege. Analyze the data from PivotTables
    created for the workbook.
    a. Open the Junior College workbook
    from the Chapter07 data folder and save
    it as Junior College Table Statistics.
    b. Create a customized list of teaching
    ranks sorted in the following order:
    instructor, assistant professor, associ-
    ate professor, full professor.
    c. Create a PivotTable with Rank Hired
    as the row label, Gender as the col-
    umn label, and count of rank hired in
    the values area.
    d. Explore the question of whether there
    is a relationship between teaching
    rank and gender. What are your hy-
    potheses? Generate the table statistics
    for this PivotTable. Which statistics
    are appropriate to use with the table?
    Is there any diffi culty with the data
    in the table? How would you correct
    these problems?
    e. Group the associate professor and full
    professor groups together and redo
    your analysis.
    f. Group the three professor ranks into
    one and redo your analysis, relating
    gender to the instructor/professor split.
    g. Write a report summarizing your re-
    sults, displaying the relevant tables
    and statistics. How do your three
    tables differ with respect to your
    conclusions? Discuss some of the
    problems one could encounter when


trying to eliminate sparse data. Which
of the three tables best describes the
data, in your opinion, and would you
conclude that there is a relationship
between teacher rank and gender?
What pieces of information is this
analysis missing?
h. Create another PivotTable with Degree
as the page fi eld, Rank Hired as the
row fi eld, and Gender as the column
fi eld. (Because you are obtaining
counts, you can use either Gender or
Rank Hired as the data fi eld.)
i. Using the drop-down arrows on the
Page fi eld button, display the table
of persons hired with a Master’s
degree.
j. Generate table statistics for this group.
Is the rank when hired independent
of gender for candidates with Master’s
degrees? Redo the analysis if neces-
sary to remove sparse cells.
k. Save your changes to the workbook
and write a report summarizing your
conclusions, including any relevant
tables and statistics.


  1. The Cold workbook contains data from
    a 1961 French study of 279 skiers dur-
    ing two 5–7 day periods. One group of
    skiers received a placebo (an ineffective
    saline solution), and another group re-
    ceived 1 gram of ascorbic acid per day.
    The study was designed to measure the
    incidence of the common cold in the
    two groups.
    a. State the null and alternative hypoth-
    eses of this study.
    b. Open the Cold workbook from the
    Chapter07 data folder and save it as
    Cold Statistics.
    c. Analyze the results of the study using
    the appropriate table statistics.
    d. Save your changes to the workbook
    and write a report summarizing your
    results.

Free download pdf