Text Size

Exercise Recommendations

Since the article excerpts included in each exercise are fictitious, you will not have to obtain the copyright clearance that is typically required when reproducing tables or other results from actual published articles. However, most exercises are inspired by actual investigations, and the references for these investigations are provided at the end of the exercises. This means that interested students can always look up real-world articles to find out what the researchers actually found.

These exercises can be used for a variety of purposes. Some examples:   

  • As in-class discussion activities
  • As homework assignments
  • As source material on which exam questions may be based

In this section, I describe how I typically use exercises of this sort. I also offer some suggestions (and warnings) for those contemplating their use.

Small-group discussion activities

In my own teaching (I teach a 400-level course on advanced statistics for undergraduates), I use these handouts almost exclusively for small-group discussion activities. I typically use an approach something like this:   

  • I lecture on a specific topic (say, basic concepts in multiple regression) for about 45 minutes.
  • I then hand out the relevant exercise (perhaps Exercise 9.3: Multiple Regression with Simultaneous Entry).
  • I lecture on the fictitious investigation described in the exercise handout for about 5 minutes, just enough so that everyone understands the basics of the investigation.
  • I assign specific items from the exercise to specific groups of students; I assign only the exercise items that are relevant to the material on which I have just lectured.
  • In their groups, the students then work on the assigned items for 10-20 minutes.
  • We get back together as a class, and the spokesperson for a given group tells the class about the group’s answers to the assigned items.
  • I provide feedback as to the rightness or wrongness of their answers (for the factual questions, anyway; some of the discussion items do not necessarily have just one correct answer).
  • I remind the students who have not had me before: Yes, this will be on the test (the students who have had me before know that everything will be on the test).

Lecturing briefly on the study described in an exercise

Many of the exercises contain fairly lengthy descriptions of fictitious investigations—descriptions that may go on for several pages. In part, this was because I wanted the exercises to contain realistically detailed descriptions—descriptions similar to those that the students would encounter in actual journal articles. These lengthy descriptions can be a problem if you plan to use the exercises as in-class discussion activities (do you really want the students sitting there for 10 minutes as they read these things?). I generally handle this by putting the exercise under the document camera and lecturing on the main points of the study. Once everyone understands the investigation’s basic features, I turn the students loose in their small groups to work on their assigned questions.

Assigning just a subset of exercise items

Most exercises have way too many items for students to complete in one setting. This means that you must always be judicious when assigning items. My standard strategy is to (a) lecture on some specific topic, and (b) then assign just the exercise items that are relevant to that specific topic. This is almost always a subset of the questions on an exercise. The upside is that a given exercise can often be spread across more than one class meeting.

 

 

Title: Advanced Statistics in Research: Reading, Understanding and Writing Up Data Analysis Results
Author: Larry Hatcher
Paperback: 644 pages
Publisher: Shadow Finch Media LLC (January 7, 2013)
Language: English
ISBN-10: 0985867000
ISBN-13: 978-0985867003
Product Dimensions: 10 x 7 x 1.5 inches
Shipping Weight: 3 pounds


If you feel this book merits a review, please take a moment to write a review.

Advanced Statistics: Reading, Understanding and Writing Up Data Analysis Results