Freshman Statistics Seminar

Week 1- Introduction to Statistics in Biology

Freshman Seminar Lesson Plan

David Schladt

Class Goals:

  • Know the goals of the course
  • Appreciate the importance of statistics in Biology
  • Understand what statistics is and why it was developed
  • Discern the difference between descriptive and inferential statistics

Class Outcomes:

By the end of the class, students will be able to:

  • State the goals of the course
  • Give a definition of statistics
  • Say who developed statistics and understand why
  • Compare descriptive and inferential statistics

Article Summary:

  • Mordiera 2006 NYT “Recommended For MDs: A Grounding In Statistics”

A new report says that doctors need at least one year of statistics to be able to properly interpret and use information from journal articles.  However, some statisticians think that doctors just need to know the “bottom line.”

Suggested Lesson Structure:

  • Syllabus and Course Goals
  • Read, answer questions about, and discuss “Recommended for MDs: a grounding in statistics”
  • Discuss
    • What is statistics?
    • Who developed it and why?
    • What can you do with statistics?
  • Background Survey: a ‘Clickers’ activity
  • Activity: Descriptive or Inferential Statistic?

Discussion Points:

  • Have the students read the article “Recommended for MDs” and answer these questions in groups of 3 or 4:
    • How much statistics training do MDs usually receive?

Six to eight weeks.

    • According to Nicholas J. Horton, how much statistics do MDs need?

At least a year.

    • What are the reasons he gives?

Can’t interpret information about up-to-date techniques

Might use information incorrectly

Might not use information at all

Doctors are at risk for manipulation by drug company pitches

May cause them to harm their patients

    • According to Stephen Lagakos, how much statistics do MDs need?

Little or none

    • What is the reason he gives?

Doctors are busy

Journals have done a good job of sorting out bad studies from good studies

  • Next discuss these questions as a class:
    • Who do you agree with and why?
    • Do you think journals do a good job of “sorting out the bad studies and keeping the good ones”?

At this point you may want to bring up Gile, Jim. 2006. Statistical flaw trips up study of bad stats. Nature. 443:379.  In the article, Steve Goodman, a medical statistician, is quoted as saying “I view most of the literature as done wrong.”

  • As a class, discuss “What is statistics?”  J. W. Kuzma (1984) Basic Statistics for the Health Sciences. Palo Alto, CA, Mayfield Publishing. “A body of techniques and procedures dealing with the collection, organization, analysis, interpretation, and presentation of information that can be stated numerically.”  Statistics has mainly been developed by agriculturalists, engineers, medical researchers, geneticists, plant and animal breeders, and economists.  For example, student’s t was developed by an engineer at Guinness, and F.A. Fisher was an animal breeder and evolutionary geneticist.  Why did they develop it?  Because they needed it!
  • Further discuss what can be done with statistics.  Another way to think about it is: What questions can’t be answered without statistics?

Active Learning Modules:

Background Survey: a ‘Clickers’ activity (EDUCAUSE Learning Initiative)

  • Ask the students multiple choice questions on any topic you what.  You may want to use the data from these questions in future activities.  Examples:
    • Number of Statistics classes taken, future career plans, number of pets, number of siblings, favorite Ninja Turtle, etc.
  • Use Clickers to collect and display the answers as bar graphs
  • Then ask “How representative do you think your answers are of all Freshman Biology students?”
  • Then collect basic demographic information about the students in the class, such as: age, gender, home country, home state, ethnicity.
  • Compare the data from the class with the statistics from all Freshman Biology students.  If the numbers are similar, your class may be a good representative sample.
  • Then ask if your students are good representatives of all college students. (Probably not.) Why or why not?  (Students’ average age is lower than all college students.)

Descriptive or Inferential Statistic?

  • First define descriptive and inferential statistics.  Descriptive statistics describe a set of measurements.  Baseball statistics are a good example.  Inferential statistics are data collected from a sample that are used to infer something about a larger population.
  • Ask what in the Clickers activity was a descriptive statistic and what was an inferential statistic.  (Data from class was descriptive, but inferential if used to describe all Freshman or all students in college.)
  • Give students a list of statistics and have them decide if the statistic is descriptive or inferential.  They can work in groups.  Examples can include:

The average family size in Austin, MN in 2000 was 2.90.

An unmarried person under 25 years old is twice as likely to have a car accident as a married person between the ages of 30 and 39.

The median household income in America in 2006 was $46,326.

The chance of receiving a full house when dealt 5 cards from a standard deck of cards is 0.144058%.

African Americans made up 12.1% of the US population in 2005.

Nolan Ryan’s career ERA (earned run average) was 3.19.

The average gestation period for the pine squirrel is 33 days.

In 2003-2004, 65% of adults 65 years of age and over reported an influenza vaccination during the preceding 12 months.

In the 2004 U.S. presidential election, George W. Bush received 50.7% of all votes.

A recent poll shows that 55% of Americans will vote for candidate A.

An unmarried person under 25 years old is twice as likely to have a car accident as a married person between the ages of 30 and 39.

The chance of receiving a full house when dealt 5 cards from a standard deck of cards is 0.144058%.

Bob has a GPA of 3.16.

The average height of 2nd graders in Mrs. Morris’s class is 1.1 meters.

Links for Additional Reading:

How to read a paper: Getting your bearings

How to read a paper: Assessing the methodological quality of published paper

How to read a paper: Statistics for the non-statistician: Different type of data need different statistical test

How to read a paper: Statistics for the non-statistician: Statistics for the non-statistician. II: “Significant” relations and their pitfalls

What your doctor should know about statistics