Developing biology: Early points of light

Last time, I wrote of problems with my approach to teaching a large-lecture non-majors class, with the hind-sight of 5 years and much reading and conversation around active learning strategies. But not everything I did then was “traditional.” My goals for that 2011 course are repeated below, and although the first two reflect very traditional approaches to teaching biology, the third is a spot of light, a sign that I was looking down a new path.

By the end of this class, you should:

  • Be familiar with the basic biological and chemical processes that make life possible
  • Be familiar with the evolutionary processes that resulted in the biological world around us
  • Be able to critically evaluate scientific information presented in the press and on-line

The goal of students learning to evaluate scientific information grew from my experiences team-teaching with Jim Bull . However, asking students to analyze texts (or “problem solve”) in prior courses often frustrated them, because most students had only experienced biology as fact-driven memorization. To support development of this skill, I taught students to use templates and “sort” items from a narrative into categories. This was the beginning of my teaching procedures of analysis as a way of understanding – although I didn’t recognize it as such. For example, the template for the scientific process looked like this:

Screen Shot 2016-06-10 at 12.49.58 PM

These templates provide students with scaffold that they could use for any text claiming to present experimental evidence. Through the course of the semester, we wandered through many examples in texts or graphs. As much as possible, the content was presented through real problems studied by real biologists. This is the kind of question I wrote:

From Discovery.com:

For a long time, doctors and diet specialists have believed that drinking more water would increase weight loss. To see if the timing of water consumption affected weight loss, researchers put more than 40 overweight and obese adults between the ages of 55 and 75 on a diet of no more than 1500 calories each day. Half the dieters were randomly assigned to drink 16 ounces of water before each meal, the others were told to drink 16 ounces of water three times a day, but not when to drink it. Twelve weeks later, the people drinking water before each meal had lost an average of 15.5 pounds while those drinking water when they pleased only lost an average of 11 pounds. The doctors concluded that for people in this age group, drinking a moderate amount of water before each meal increased the rate of weight loss.

Which component of the scientific method is absent?

a) Hypothesis

b) Protocol

c) Evaluation

d) Revision

e) All are present (none are absent)

Which of the following hypotheses is rejected by these results?

a) Eating more than 1500 calories each day can be countered by increased water consumption.

b) Among younger people, increased water consumption does not lead to increased weight loss.

c) Drinking 16 ounces of water before each meal causes increased weight loss among people who are dieting.

d) The timing of water consumption has no effect on its role in weight management in older people.

Why did I use real data? Several reasons, but most importantly because the unsanitized data presented students with a glimpse of the complexity found in the real world and how scientists work with, around, or through that complexity. It also gave me the opportunity to uncover and work with students’ misunderstandings about the nature of science as a human endeavor. Frankly, the sanitized experiments look prescient and self-fulfilling, lacking the sweat and frustration that are the hallmark of cutting edge research.

As an example of this, we can contrast the sanitized presentation and the real data from  Griffith’s classic experiment on transformation of bacteria. In this experiment, Griffith studied the impact of variation using two strains of a bacteria. One strain, called “smooth” because colonies in a Petri dish look smooth, causes deadly pneumonia; the other strain, called “rough” because colonies look rough, does not. (More about these two strains, the biology of their appearance, and their ability to cause disease can be found here ). The classic presentation of Griffith’s research into the dynamic relationship between these two strains of bacteria, found widely in introductory biology and genetics textbooks, looks like this:

 

450px-Griffith_experiment.svg
Injecting a mouse with the rough strain does not cause disease, while injecting with the smooth strain does. Injecting dead smooth strain  does not cause disease, but injecting a mixture of living rough strain and dead smooth strain does. From Wikipedia

 

You can, however, read Griffith’s original  paper (1928 Journal of Hygiene 27 p 8-55), and it is fascinating – in no small part because Griffith’s personality and thought processes emerge in his writing. He describes everything, from his initial observations and pilot studies through  the classic experiment so succinctly summarized in that textbook figure.

Importantly, Griffith actually did many replicates of his experiment using different genetic strains of smooth and rough bacteria. For my class, I pooled data across several replicates to produce a table that looked like this:

Treatment Sample size Prediction Results
Inject living R 6 Mice survive; recover R 4 mice survive; recover R. 2 mice die; recover S
Inject dead S 26 Mice survive, no bacteria recovered 26 mice survive, no bacteria recovered
Inject living R and dead S 71 Mice die; recover S It’s complicated!

The actual results from injecting living R strain together with dead S strain was not 100% mortality and recovery of S strain bacteria from the dead, as reported in textbooks. Instead, it is complicated – Griffith’s data looked like this:

Screen Shot 2016-06-10 at 10.38.59 AM
The percent of mice producing each type of bacteria; S-infected mice died, R-infected and uninfected mice survived.

 

It took time to walk through the experiment with this level of detail. I started with the background observations and development of the original hypothesis of transformation. We talked about what variables should be manipulated, what should be measured, and what should be controlled for. We talked about the idea of distribution of outcomes, and possible causes of the variation.

What did we gain by talking about real research and real data? First, exposure to the role of variation in biology – something that arguably makes biology a very difficult science to understand. Some of the variation in Griffith’s results reflects my pooling across his many replicates, which were done with different strains of bacteria, some of it is actual variation for a single strain. Second, the appearance of probabilistic results prompted conversation around the need for replication (one prompt I used was “if he injected just one mouse with living R strain, what is the likelihood that mouse would die?”). Taken together, these two points created a starting point for conversations around why sometimes it seems that scientists “change their minds” in the media.

Why is this important? Recall, I was teaching non science majors in this class, where “working scientist” stories are often relegated to side boxes in textbooks and “if you are interested” extra readings and links. Moreover,  working-scientist stories are often already sanitized (no mistakes! no wrong turns! no variation!), may seem very tangential to student lives, or may not actually mention the scientists by name (for example Audesirk et al. 8th ed scientific inquiry stories).

These sanitized presentations of how science happens reinforce the perception of the scientist as the isolated genius, the hypothesis as somehow given, the protocol as perfect, and the results as inevitable. How could a “regular person” ever hope to achieve this, or understand how it happens? If our goal is to convince our students that they can understand and undertake inquiry, we need to show them that the process is messy – mistakes are made, protocols don’t work, hypotheses are rejected and need to be rethought. Exposure to real studies necessarily reduces the content coverage we achieve in introductory courses. But in my opinion, this is a small price to pay for increased understanding and self-efficacy in science.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s