itrn 501: fall 2008 methods of analysis for international commerce and policy class 2 instructor:...
TRANSCRIPT
![Page 1: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/1.jpg)
ITRN 501: Fall 2008
Methods of Analysis for International Commerce
and PolicyClass 2
Instructor: Danilo [email protected]
![Page 2: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/2.jpg)
Objectives of this classStudents will:• be proficient in the most basic quantitative methods
used in international commerce policy.
• have experience retrieving and formatting quantitative data from standard sources.
• be familiar with multivariate analysis methods.
• understand some of the most prevalent practical problems and ethical issues confronting policy analysis.
• complete a project drawing on their knowledge of these elements.
![Page 3: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/3.jpg)
What are data?
• A representation of facts, concepts, or instructions in a formalized manner suitable for communication, interpretation, or processing by humans …
• Characteristics of data determine the possible methods of analysis
![Page 4: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/4.jpg)
The subjective/objectivenormative/positive debate
In popular usage: • Objective matters can be observed and
quantified and all must reach the same basic result in assessing them.
• Subjective matters are open to individual interpretation.
• Positive statement - a statement of “fact” without indication of approval.
• Normative statement - expresses whether a situation is desirable or undesirable.
![Page 5: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/5.jpg)
The qualitative/quantitative debate
• Quantitative data can be counted and the results of statistical analysis are meaningful.– Basic interpretation is clear, i.e. x>y.
• Qualitative data are meaningful to humans but can not be counted or manipulated with statistical methods.– Researcher/reader must be relied on for
basic interpretation.
![Page 6: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/6.jpg)
The practical synthesis• The debate is limiting to the policy analyst. Data and
methods should be assessed as they are useful and necessary to address a problem.– There is plenty that can be subjective, normative and
qualitative in quantitative analysis.
– Sometimes qualitative data are the best or only data there are
• Mixed methods often lead to better questions and stronger, more persuasive results, reaching broader audiences.– Case studies and anecdotes can motivate, explain, support, or
raise questions about quantitative results.
• With coding, qualitative data can be introduced to quantitative models.
![Page 7: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/7.jpg)
Types of data used in this class (scales)
• Numeric variables – – Interval data have meaningful intervals
between measurements, but there is no true starting point (zero).
• 20 C is twice 10 C but 68 F is not twice 50 F
– Ratio data have the highest level of measurement. Ratios between measurements as well as intervals are meaningful because there is a starting point (zero).
– Basic policy research mostly assumes ratio scales.
![Page 8: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/8.jpg)
Types of data used in this class (scales) cont.
Ordered Categorical Variables – or ordinal data allow the ranking of the data, e.g. bigger/smaller, healthier/less healthy etc., but the interval is meaningless
• Unordered Categorical Variables – or nominal data are categorical data where the order of the categories is arbitrary, e.g. race, religion, colors.
![Page 9: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/9.jpg)
Types of statistical data (scales)
• Discrete Variable – Has a limited number of known values, e.g. number of automobiles imported by Ghana can not be 100,000.2.
• Continuous Variable – Can take any value, weight tends to have this quality and currency does to a great extent. Ghana could import 1,000,000.25713 KG worth of automobiles, – even if it is not likely significant at this level of
precision.
![Page 10: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/10.jpg)
The use of statistical methods to analyze data does not
(necessarily) make a study more “scientific”, “rigorous”, or
“objective.”
1) The wrong method
2) The wrong data
3) The wrong question
4) Just wrong (error)
![Page 11: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/11.jpg)
Error in data and analysis• Random error
– Sampling error– Random misclassification
• Systematic error/bias– Systematic non-random deviation from the true
values.– Can be conscious or unconscious.
– Need not be “on purpose.”
– Bias creates an association that is untrue.– Confounding error creates an association that
is true but potentially misleading.
![Page 12: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/12.jpg)
Ideally, problem determines methods and data, and these in turn your conclusions…– You should not assemble data to prove your point.
(Sometimes we can be selective to make a point.)– Method choice or data availability should not
determine problem definition, • i.e. if you have a hammer you should not make
every problem a nail.
(We are unaware of all possibilities and they are not always at our disposal.)
In sum, we try not to use statistics as a drunk might use a street lamp:– For support rather than illumination, or – To decide where (or what) to look for.
![Page 13: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/13.jpg)
Thinking in models
• What is a model?– Explains which elements relate to each
other and how.– Describing Relationships in a model
• Covariation – move in the same direction– Direct or Positive – Inverse or Negative– Nonlinear
• False of spurious– Control (confounding) variables
• Are you looking for the best model or testing someone else’s?
![Page 14: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/14.jpg)
Developing models
• Where does a model come from?– From your own assessment and
observation of the problem, or from talking to others.
– From the literature.• Elements others include or consider important• Definitions of these elements • Descriptions of the “expected” relationships
among variables• Results and explanations• Sources and strategies for data• Suggestions of models or variations to be
tested in the future
![Page 15: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/15.jpg)
Types of Models
1. Symbolic• Economic growth is a function of
changes to the amount of capital (K) and changes to the amount of Labor (L).
• G=f(K,L)
• G=α+β1K+β1L+e
2. Schematic Capital
Labor
Econ Growth
![Page 16: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/16.jpg)
The importance of writing• Policy writing is a fundamental form of analysis:
– Written results must “track” and be accessible. • If it does not make written sense, and the argumentation
does not follow, the analysis is suspect.
– Writing helps the researcher and not just the reader understand the results.
– Results without a well written analysis will generally have less policy influence.
• Bad results with good writing often have a greater impact than they deserve.
![Page 17: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/17.jpg)
The importance of critical thinking and logic
• Received wisdom is not always right…• But if want to say that it isn’t you need to
recognize it, and address its failings.• Familiarize yourself with common fallacies
• http://www.unc.edu/depts/wcweb/handouts/fallacies.html• http://www.nobeliefs.com/fallacies.htm• http://www.nizkor.org/features/fallacies/
– Hasty generalization– Unrepresentative sample– Post Hoc– Straw man– Category errors– Non sequitor
![Page 18: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/18.jpg)
Tables and figures
• Must also include anything necessary for proper interpretation. Exhibits must be able to stand ALONE.• Titles – tells reader what is going on,
what they are looking at, may provide some interpretation.
• All relevant data, no irrelevant data must be included.
• Clear labels titling data and units• Sources
![Page 19: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/19.jpg)
Tables
• Tables are used to present many data series or variables or when details are important.
• Columns should be fewer than rows in most instances.
• Nested tables, crosstabs etc.
![Page 20: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/20.jpg)
Source: World Bank (2006) Moldova Poverty Update
![Page 21: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/21.jpg)
Line graph
• Often best to show change over a series of points in time, or any continuous change (i.e. income distribution)
• X axis (time) series variable
• Y axis variable of interest
![Page 22: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/22.jpg)
Source: World Bank (2006) Moldova Poverty Update
![Page 23: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/23.jpg)
Bar graph
• Can be used with just two data points• More visually striking when fewer
data points are expressed. • For comparisons of multiple
observations over a few years it can overcome the spaghetti problem of line charts.
• Can be combined with line charts to good effect.
![Page 24: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/24.jpg)
Source: World Bank (2006) Moldova Poverty Update
![Page 25: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/25.jpg)
Source: World Bank (2006) Moldova Poverty Update
![Page 26: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/26.jpg)
Pie chart
• Used to show proportions and shares at a point in time
• Must add up to a meaningful total• Often used for comparisons when
other charts would be preferable.
![Page 27: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/27.jpg)
Figure 2.1 Types of Drugs Used by Past Month Illicit Drug Users Aged 12 or Older: 2003
D
Source: US DHHS (2004) 2003 National Survey on Drug Use & Health: Results
![Page 28: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/28.jpg)
Shares in bars: better for comparison
![Page 29: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/29.jpg)
Issues and tricks
• Scale and origin• Using indexes to compare variables with
different scales.• Normalize by a variable such as population.• Show only the most important relationships.
– Provide full data in appendix tables
• Titles can lead reader as long as subtitles, and all other required information are clear and complete
![Page 30: ITRN 501: Fall 2008 Methods of Analysis for International Commerce and Policy Class 2 Instructor: Danilo Pelletiere dpelleti@gmu.edu](https://reader036.vdocuments.us/reader036/viewer/2022081515/56649e2a5503460f94b17d27/html5/thumbnails/30.jpg)
The final project
• What are others saying about trade and your country?
• What is your model?– What is happening and why?
• Do you have the data you need?– Can you get them?– What do you think they say?
• Is the data ready for presentation?• Start writing and be ready to reiterate
these steps.