Objective: Identify and use the four principles of experimental design. HW: Read pp. 290-293 and complete exercises 5.31, 5.32, 5.33

Then read pp. 294-298 and complete exercise 5.38 (which is an extension of 5.32).

Do Now: Answer the following questions in your notebook. If you

don’t know, simply leave the question blank:

(a) What did you eat last night for dinner two nights ago?

(b) What color shirt did you wear yesterday?

(c) What time did you wake up yesterday morning?

(a) How confident are you that your answers to the above questions are accurate?

Observational Studies

• Observes people “in the wild.”

• Researchers don’t assign choices or manipulate subjects. They simply observe them.

• What do you think the difference between a retrospective and prospective study are?

• Retrospective data identifies subjects and then looks at data from their past while prospective data identifies subjects and then collects data as the events unfold.

• Observational studies may identify important variables related to the outcome we are interested in, but there is no guarantee that we have found the right or most important related variables.

Is it ever possible to prove a cause

and effect relationship?

Yes! But only if we run an EXPERIMENT.

Does taking vitamin C reduce the chance of getting a cold?

• The individuals on whom or which we experiment are called experimental units. Human experimental units are called “subjects” or “participants.”

Does taking vitamin C reduce the chance of getting a cold?

Our experimental units are probably humans

in this scenario.

• Experiments study the relationship between two or more variables.

• An experimenter must identify one explanatory variable, called a factor, to manipulate and at least one response variable to measure.

• An experimenter manipulates the factors to control the details of possible treatments.

Does taking vitamin C reduce the chance of getting a cold?

Explanatory Variable / Factor (to

manipulate):

Response Variable (to measure):

• The specific levels that the experimenter chooses for a factor are called the levels of a factor.

• The combination of specific levels from all the factors that an experimental unit receives is known as its treatment.

Does taking vitamin C reduce the chance of getting a cold?

Let’s choose levels for our factor:

Response Variable

• How will you measure your response variable?

• An experiment requires a random assignment of subjects to treatments.

DIAGRAMS

Group 1 Treatment 1

Random allocation Compare

Group 2 Treatment 2

The Four Principles of Experimental Design

CONTROL

• Control sources of variation other than the factors you are testing by making conditions as similar as possible for all treatment groups.

RANDOMIZE

• Randomization allows us to equalize the effects of unknown or uncontrollable sources of variation.

REPLICATE

• Repeat the experiment with different subjects. Only with replication, can we estimate the variability in responses.

• If we experiment on a single subject, the outcome is an anecdote.

Because I know you love them:

REPLICATE

• We also can talk about replication of an entire experiment at different levels (in other parts of the country, with people of different ages, at different times of year).

BLOCK

• This is the only one that is not required for every experiment. This is analogous to stratifying for sampling design in surveys.

• We can group similar individuals together and then randomize within each “block.”

Designing Experiments

• An ad for OptiGro plant fertilizer claims that with this product, you will grow “juicier, tastier” tomatoes. You’d like to taste this claim, and wonder whether you might be able to get away with half the specified dose. How can you set up an experiment to test this claim?

1) State what you want to know: I want to know if OptiGro makes tomato plants tastier or juicier than plants without fertilizer.

2) Specify the response variable: The level of juicy and tastiness in the tomatoes.

3) Specify the factor levels and treatments I will grow tomatoes at three different levels- some with no

fertilizer, some with half the specified amount, and some with the full dose. These are the three treatments.

4) Specify the experimental units: 24 lovely tomatoes (possibly organic and of the same variety) from the same store.

Observe the principles of experimental design:

-Control any sources of variability you know of and can control

-Randomly assign experimental units to treatments to equalize the effects of unknown or uncontrollable variation

-Specify how the random numbers needed for randomization will be obtained.

-Replicate results by placing more than one plant in each treatment group.

Grow tomato plants in the same soil / plot of land, take into account environmental things like trees, etc. Take care of them consistently (same amount of watering, pesticides, etc)…

I will use my calculator to randomly select tomatoes to put into 3 different groups.

24 Tomatoes – put 8 in each treatment group

Other Details

• Specify other experiment details. You must give enough details so that another experimenter could exactly replicate your experiment. It’s generally better to include details that seem irrelevant than to leave out matters that could turn out to be important.

• Specify how to measure the response.