Statistical Investigation
From MM*Stat International
English |
Português |
Français |
Español |
Italiano |
Nederlands |
Conducting a Statistical Investigation
Statistical investigations often involve the following steps:
- Obtaining data
- Primary data: data collected by the scientist
- Surveys:
- observing all members of the population (census) or taking a sample (sample survey)
- documentation of data via questionnaires, protocols etc.
- Experiments: actively controlling variables to capture their impact on other variables
- Surveys:
- Secondary data: using readily available data, either from internal or external sources.
- Primary data: data collected by the scientist
- Analysis: applying statistical tools
Sources of Economic Data
- Public Statistics
- Private Statistics
- International Organizations
The following figure illustrates the sequence of steps in a statistical investigation:
The official body engaged in collecting and publishing Berlin-specific data is the Statistische Landesamt Berlin. For example, statistics on such disparate subjects as the animal populations of Berlin’s three major zoos and voter participation in general elections, are available.
Animals in Berlin, 1998 | Zoo and Aquarium | Safari |
---|---|---|
Mammals | ||
Total population | 1 357 | 1 861 |
Species | 247 | 214 |
Birds | ||
Total population | 2 333 | 2 548 |
Species | 500 | 408 |
Snakes | ||
Total population | 461 | 723 |
Species | 82 | 129 |
Lizards | ||
Total population | 412 | 87 |
Species | 31 | 5 |
Fish | ||
Total population | 4 007 | 1 680 |
Species | 360 | 136 |
Invertebrate | ||
Total population | 6 298 | 1 368 |
Species | 181 | 72 |
Visitors | 2 597 589 | 1 071 207 |
Elections into the 14th German “Bundestag” in Berlin (27.9.1998). The map displays the election participation in districts of Berlin.
A common objective of economic policy is to reduce the overall duration of unemployment in the economy. An important theoretical question is, to what extent can the level of unemployment benefits account for variation in unemployment duration In order to make this question suitable for a statistical investigation, the variables must be translated into directly observable quantities (For example, the number of individuals who are registered as unemployed is a quantity available from government statistics. While this may not include everyone who would like to be working, it is usually used as the unemployment variable in statistical analyses). By examining government unemployment benefit payments in different countries, we can try to infer whether more or less generous policies have an impact on the unemployment rate. Prior to further investigation, the collected raw data must be organized in a fashion suitable for the statistical methods to be performed upon them. Exploring the data for extractable information and presenting the results in an accessible fashion by means of statistical tools lies at the heart of statistical investigation. In interpreting quantitative statistical information, keys to an answer to the initial scientific questions are sought. Analogous to the general scientific process, conclusions reached in the course of statistical interpretation frequently give rise to further propositions—triggering the next iteration of the statistical process.