Streamlining Python Development: A Guide to a Modern Project Setup
Quantitative Methods for Lawyers - Class #4 - Research Design Part IV - Professor Daniel Martin Katz
1. Quantitative
Methods
for
Lawyers Research Design - Part IV
Class #4
@ computational
computationallegalstudies.com
professor daniel martin katz danielmartinkatz.com
lexpredict.com slideshare.net/DanielKatz
2. In our last session, we were
discussing randomized control trials
12. Random Assignment
P r o b a b i l i t y o f B e i n g i n
Treatment or Control Group
Should Be Equal
Composition of the Treatment or
Control Group Should Be
Similar
Under Ideal Conditions this
Eliminates other Confounds that
could undermine Validity
13. Would like to overall subject
group to mirror the population
of interest
Example: If we are interested in
studying juveniles than the
composition of bot h our
treatment and control groups
should be juveniles
Representativeness
14. Experimental Control
Classic Example is Medical Trial Involving a New Drug
Experimental Group
Given the New Drug
Control Group
Given the Sugar Pill
How would Double Blind work in this context?
What is a Placebo Effect?
15. Experimental Manipulation
Under Ideal Conditions this would be the only
difference between treatment and control group
Experimental Manipulation
and Factorial Design
Watch out for too
many Manipulations
at one time
17. Concept: A variable is an attribute which
describes a part of the makeup of an individual.
Examples are gender, age, employment status,
income level, race, or education level.
18. Studies are usually designed to collect and
then compute the distribution and variation
between and among the variables.
19. It should be noted that a variable, by
definition, must possess variation; if all of the
studied population have the same attribute,
for example they are all employed, that
attribute is a constant rather than a variable.
20. There are different types of variables.
One important division is between
independent variables and
dependent variables.
21. Independent variables act as the
potential cause. They influence or
predict an outcome from the dependent
variable. They are the X’s on the right
side of the equation.
22. Dependent variables act as the effect
(or potential effect). They may
change because of the influence of the
independent variable. This is the Y on
the left hand side of the equation.
25. Nominal variables are variables that have
two or more categories but which do not
have an intrinsic order. For example, a real
estate agent could classify their types of
property into distinct categories such as
houses, condos, co-ops or bungalows.
26. Dichotomous variables are nominal variables
which have only two categories or levels. For
example, we could categorize somebody as
either Treated or Not Treated as either "Yes"
or “No”. In the real estate agent example, if
type of property had been classified as either
residential or commercial then "type of
property" would be a dichotomous variable.
27. Ordinal variables are variables that have two
or more categories just like nominal variables
only the categories can also be ordered or
ranked. Large, Medium, Small, etc.
28. Describe some variables could
that could predict/determine
the price of a house?
How Are They Coded?
29.
30. School Quality
New or Used
Pool
Garage
Distance from City Center
... etc.
BedRooms
BathRooms
Square Feet
Lot Size
Age of House
Crime Rate
32. Please note that “bias” in research terms is
different.
In normal language, bias is a prejudicial look at
someone or something.
In research, bias is an action or inaction which can
skew the outcome.
It does not have to be intentionally done.
Bias in Scientific Study
37. Sometimes the statistical test shows a clear and
significant relationship called a correlation between
two variables.
There is a tendency to then conclude that the
correlation shows causation. It may (or may not).
It could have nothing to do with causation or it could
only have an indirect affect on the causation.
38.
39. Daniel Martin Katz
@ computational
computationallegalstudies.com
lexpredict.com
danielmartinkatz.com
illinois tech - chicago kent college of law@