Sep 30, 2009

What do the values in time method mean?

  • Longitudinal - Data collected repeatedly over time to study change in a population.
  • Longitudinal:Cohort/Event-based - Data collected over time about a group of individuals that are connected in some way or have shared some significant experience within a given period. Examples: birth, disease, education, employment, family formation.
  • Longitudinal:Trend/Repeated Cross-section - Studies different samples /groups of people from the same population at several points in time; conclusions are drawn for the population. Examples: public opinion polls, elections studies, etc.
  • Longitudinal:Panel - Data collected over time from, or about, the same sample of respondents.
  • Longitudinal:Panel:Continuous - Reports from the panel are collected on a regular basis.
  • Longitudinal:Panel:Interval - Measurements are taken only when information is needed.
  • Time Series - Data are collected repeatedly over time to study change in observations.
  • Time Series:Continuous - Phenomena are measured at every instant of time. Examples: lie detectors, electrocardiograms, etc.
  • Time Series:Discrete - Measurements are taken at (usually regularly) spaced intervals. Examples: macroeconomics (weekly share prices, monthly profits, sales, etc.), meteorology (daily rainfall, hourly temperature, etc.), sociology (crime figures, employment figures, etc.).
  • Cross-sectional - Data about a population are obtained only once.
  • Cross-sectional ad-hoc follow-up - Data collected at one point in time to complete information collected in a previous cross-sectional study; the decision to collect follow-up data is not included in the study design.

Please note that ICPSR has not yet retroactively applied the time method field to all 8000+ studies we archive.

Sep 1, 2009

How do I use the Recode Syntax analysis tool?

The Recode Syntax tool is found with the online analysis utilities. Users must first define a variable name to generate the recode. In this example, we are choosing to create a recode that cross-classifies race and age. It creates a dummy variable that takes on the value of 1 if the respondent identifies themselves as Black and over 18 years of age.

Screen Shot

Users can specify the content of the recode with a series of dropdown variables, logical operators, and mathematical functions. The choices are available in the dropdown values linking the expressions.

Screen Shot

To add additional variables, click the "Add" button on the right side of the screen. This will continue to add variables with which the user can build recodes. Below, race is added to age to generate the recode of interest. Users must complete both sides of the recode, in order to create two sets of recode statements.

Screen Shot

Users can continue to add recodes to the syntax by clicking the "Recode Syntax" button.

Users must then choose a statistical package of choice and click the "Submit Code" button. The code will then appear in the window and can either be printed or downloaded. Again, this tool does not generate recodes but rather the software code needed to generate them. They must be run in the appropriate software using the dataset from which they are generated.

Screen Shot

Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial 3.0 United States License.