Quantitative Methods and Computer Applications for Economic and Social History QM1

1
Quantitative Methods and Computer Applications
for Economic and Social History- QM1 -

• Dr Alexander Moradi
• GPRG/CSAE Nuffield College
• Dept. of Economics, University of Oxford
• Email alexander.moradi_at_economics.ox.ac.uk

2
Contents

• Introduction
• STATA Basics
• Descriptive Statistics
• Correlation
• Simple Linear Regression
• Statistical Significance
• Multiple Linear Regression
• Dummy Variables

3
1. Aim of the Course

• Provide an introduction to elementary
  quantitative methods and statistics, covering
  some of the techniques most widely used in
  research in the historical and social sciences
• To promote the understanding of quantitative
  techniques by practising the techniques on real
  data
• Introduce the statistical software package STATA
  and show how the relevant statistical
  applications can be performed
• Discovering applications for your own research
  projects

4
1.Advantages of Quantitative Methods

• How representative or reliable are observed
  patterns?
• Exceptions are taken as such
• Observed patterns may be just random
• If many factors determine an outcome, the
  relative importance of the factors can be
  determined
• Graphs, tables, cartography can be very
  persuasive tools, they can also give valuable
  hints and raise new questions
• Its good for the job market

5
1.Cliometrics

• Empirical studies based on facts or experiments
  in Economics/ History/ Social Sciences based on
  observations generated by the real world
• Empirical studies test hypotheses by using
  statistical techniques on data
• Cliometrics refers to the systematic use of
  economic theory and quantitative techniques to
  study economic history

6
1. Why STATA?

• STATA is a good option no need to switch to
  another software when you do a more ambitious
  study
• No software can substitute an analyst she
  plays a crucial role in choosing an appropriate
  test strategy and interpreting the results!

7
Useful Literature

• Feinstein, C. H. and M. Thomas (2002). Making
  History Count A Primer in Quantitative Methods
  for Historians. Cambridge Cambridge University
  Press.
• Blalock, H. M. (1979). Social Statistics. New
  York McGraw-Hill
• Huff, D. (1991). How to Lie with Statistics.
  London Penguin Books
• Wonnacott, T. H. and R. J. Wonnacott (1990).
  Introductory Statistics. New York Chichester
  Wiley
• Wooldridge, J. M. (2006). Introductory
  Econometrics A Modern Approach. Ohio South
  Western College Publishing
• Stata Corporation (2005). Stata User's Guide
  Release 9. College Station, Tex. Stata Press

8
Course Assessment

• Satisfactory completion of the weekly homework
  assignments
• At the end of the course there will be a simple
  take-away examination to test understanding of
  the various concepts and procedures covered
  during the course.
• Two typewritten copies of the completed
  assignment must be submitted by noon on Monday of
  Second Week of Hilary Term to the Chairman of
  Examiners for the MPhil in Economic and Social
  History, c/o The Clerk of the Schools,
  Examination Schools, High Street, Oxford OX1 4BG
• It is necessary to pass this Qualifying Test in
  order to proceed to the main examination. The
  results count for 10 percent of the final MSc
  degree mark, and 7 percent for the MPhil degree
  mark

9

• You can download slides and data from
• http//www.economics.ox.ac.uk/Members/
• alexander.moradi/teaching.html
• If you have difficulties to follow, please give a
  signal and ask. It takes some time to get used to
  the new terminology

10

• Any questions or comments so far?
• Short introduction of participants
• Whats your background and how much do you know
  already about statistics and quantitative
  methods?
• Are you planning to use quantitative methods in
  your research?

11
QM1 Week 0 Getting started

• Dr Alexander Moradi
• GPRG/CSAE Nuffield College
• Dept. of Economics, University of Oxford
• Email alexander.moradi_at_economics.ox.ac.uk

12
2.The Structure of Data

• Data sets (Random) samples of individuals,
  households, firms, countries etc. or
  comprehensive surveys/complete population
• Cross-sectional data Cross-section at a given
  point in time
• Time series data Observations on one unit over
  time
• Combination of cross-section and time series
• Pooled cross sections cross-sections at
  different point in times, cross-sectional units
  change
• Panel/ Longitudinal same cross-sectional units
  are followed over a given time period

13
2. Cross-sectional data

• Example Real wages across Europe in 1750

Source Allen (2001)
14
2. Time-series variables

• Example Real wages in London, 1500-1800

Source Allen (2001)
15
2. Panel data

• Example Real wages across Europe, 1500-1800

Source Allen (2001)
16
2. Variables, Cases, Values

• Cases are the basic unit of observation, and they
  can be individuals, households, firms, towns,
  countries, etc.
• Variable Specific characteristic of a number of
  observations that vary from observation to
  observation
• Value The specific characteristic that an
  observation has
• Example English Poor Law data set
• 311 English parishes Cases? Variables? Values?
• Relief expenditure of each parish Cases?
  Variables? Values?
• 20.4, 16.3 Cases? Variables? Values?

17
2.Levels of Measurement

• Nominal measurement
• Each value defines a distinct category but gives
  no information other than the label or name
  (hence nominal level)
• Example
• Cases Migrants
• Variables Birthplace of migrants
• Values London, Oxford, London
• Ordinal measurement
• This applies when it is possible to order or rank
  all the categories according to some criterion
  without being able to specify the exact size of
  the interval between any two categories
• Example
• Cases Migrants
• Variables Skill level of migrants
• Values unskilled, semi-skilled, skilled

18
2.Levels of Measurement

• Interval or ratio measurement
• These measurements have all the properties of an
  ordinal scale. In addition, it is now possible to
  measure the exact distance between any pair of
  values
• Values are either continuous or discrete
• Example 1
• Cases Migrants
• Variables Number of children
• Values 0, 2, 7
• Discrete variable! Values can take only a number
  of pre-determined values
• Example 2
• Cases Migrants
• Variables Height of migrants
• Values 1.65m, 1.72m, 1.68m
• Continuous variable is measured in units that
  can (theoretically) be reduced in size to an
  infinite degree

19
2.Populations and Samples

• Population
• Refers to all possible observations. In the Poor
  Law example that would be all parishes in the
  whole country
• Sample
• Is a subset of the population
• Either because not all cases are available or
  because it would take too much time to study all
  cases
• This is not a big problem, as there are tools to
  make valid inferences from a sample
• Crucial Is it a random sample or not?

20
2.STATA Basics

• Various data formats and extensions
• .xls Excel
• .dbf dBase
• .sav SPSS
• .dta STATA
• STATA can import Excel/text files
• Stat/Transfer Data conversion package software
  that facilitates file transfer between a number
  of formats
• To keep the overview, it is important to record
  coding, calculations, manipulations, e.g. in a
  word processor
• Never overwrite your original data set!

21
2.File types in STATA

• Data files .dta Contains the data
• Graphs .gph Stata graphs Stata can only open
  the files with the .gph extension, other packages
  cannot read this format
• Log files .scml Records the output of STATA
  sessions (everything you saw in the results
  window) does not record the graphs
• Do-files .do Rather than typing commands in the
  command window, you can create a Do-file
  containing a series of commands and executed by
  STATA
• Ado-files .ado Defines a STATA command when you
  instruct STATA to execute a command STATA runs an
  ado-file useful for programming dont change
  them!

22
2.Exercise Getting Started

• Data set Depression.dta
• The data set includes macroeconomic data like
  industrial production, prices, etc. for a number
  of countries during the Great Depression 1929-36
• lthttp//www.library.northwestern.edu/govpub/collec
  tions/league/stat.htmlgt
• Exercises
• Open the file and use the Data Editor to
  understand the data structure
• What do the columns and rows mean?
• What structure does the data have?
• Change the name of variables and labels
• Import the Exel file weimar_election.xls into
  STATA
• Save the Excel File as CSV (comma delimited)
• In STATA File/Import/ASCII data created by a
  spreadsheet
• What structure does the weimar_election data
  have? Aggregate the data, average the values of
  the regions
• In-putting data

23
2. STATA commands