paper exercise 2 1



Restate your observation (already done in first assignment). If you wish or feel the need to do so,

you can pick a new observation and start over. If you do this, please refer to the first assignment in order to

remind you what is required. If you need to revise anything that you, please do so.


Restate your three models (already done in first assignment). Many of you may have to come up

with new models. Also, if it was stated poorly in the first exercise, fix it now. As always, remember

PROCESS. Then make sure you have followed the other rules of thumb: Generality and interesting

implications. Also, remember this isn’t a list itemizing different reasons for something, but rather models

that are explaining how or why something happens. Ideally, for this assignment, they should be competing

models where only one of them can be true. The format should be the same as in the first homework

assignment. The first sentence is a simple sentence stating the model. The next couple of sentences can

explain how the model relates to the observation. The model should be separate from the observation.

Most people have a tendency to state a model too specific to the observation. Watch out for this.


State two interesting implications (hypotheses) for each model. Remember to refer to Lave and

March for what makes a good model. There are new concepts, also: Simplicity and Fertility. If you are

not able to come up with a wide variety of implications and hypotheses, it may be because your model isn’t

fertile. Remember the rules of thumb (Generality and Interesting Implications). A few sentences should

suffice for each implication.


Describe natural experiments that will discriminate between the models. We call any experiment

that distinguishes between two models a critical experiment. So, these should be natural experiments that

are also critical experiments. (We will refine this more in the next assignment.) There need to be at least

two experiments (if it is well done). Three would probably be better. Do not combine this section with

section 3 (above). It is a distinct section although many of you will find it repetitive. Depending on your

observation and your models, the natural experiments may be very similar to a re-stating of your

implications. If this occurs, don’t panic. Likewise, don’t expect or anticipate this to happen.

The following four points are guidelines/requirements.


The experiments must be critical experiments.


You must have enough critical experiments to distinguish all of your models from each

other. Remember, if an experiment predicts the same result for each model, then it is not a good



All of the experiments must be natural experiments


Make sure that you specify which model is supported by a particular experimental

outcome. And make sure that for each experiment you discuss more than one outcome.

In addition, you will have to operationalize all of the variables used in the experiments, and you will have

to indicate which variables are dependent variables and which variables are independent variables.

Thus, your assignment will look like this:

Section One (old/revised): A description of your observation.

Section Two (old/revised): Stating your first model and explaining how it relates to your observation.

Section Three (new): State the two implications for your first model. Discuss each implication separately.

Use two paragraphs, one for each implication.

Section Four (old/revised): Stating your second model and explaining how it relates to your observation.

Section Five (new): State the two implications for your second model. Discuss each implication

separately. Use two paragraphs, one for each implication.

Section Six (old/revised): Stating your third model and explaining how it relates to your observation.

Section Seven (new): State the two implications for your third model. Discuss each implication separately.

Use two paragraphs, one for each implication.

Section Eight (new): Describe one of your critical experiments. (Remember it has to be a natural

experiment as well.) State what the experiment is. State which two models it is distinguishing between.

And state what you would expect to see if one of the models is true, and state what you would expect to see

if the other model is true.

The description of the first critical experiment should look something like


Describe the experiment.

b) State which are the independent and dependent variables.


State how you are going to operationalize the variables.

d) Describe the outcomes and state which model each outcome supports

Section Nine (new): Repeat Section Eight with another critical experiment.

Section Ten (new): Repeat yet again with yet another Critical experiment. (Remember, unless you can

distinguish all three of you models from each other with only two critical experiments, it is strongly

recommended that you use three critical experiments to successfully distinguish all of you models from

each other.) Note: If your first experiment distinguishes Models 1 and 3, your other experiments cannot all

distinguish Models 1 and 3 and never discuss Model 2. I assume this is obvious.

You should separate the sections. Do not merely write one long paragraph. This isn’t a literary society.

We want to easily see you points.

Realize that we are expecting more at this point. Just because you received nine or ten points on your

previous assignment does not mean that you will earn as many points for the same work here.