Evalguide -

Depending to the choice of data collection method and question, the data can be analysed using most different statistical or content analysis methods. We cannot present all of these analysis methods in detail in this handbook, therefore we will confine ourselves to basic information and provide sample analyses for some evaluation questions.

Analysis of criteria catalogue data

The analysis of criteria catalogue data is determined based on an analysis mode defined within the catalogue itself. This should be followed, otherwise false results can be obtained.

Analysis of network analysis data

Each project within a program constitutes a node in the network of which the interconnectedness between the nodes is to be shown.

Analysis example

One option is to represent the relationships between the projects, for which data has been collected using the questionnaire on networking (see page on Network analysis), simply as connections between the nodes (cf. Netzwerkdarstellung 1 [network diagram 1] (pdf, 276 kB, in German); with a diagram for each dimension: Orientation, familiarity, etc.). The dimensions orientation / familiarity and input / output can also be visualised as part of a coordinate system: Netzwerkdarstellung 2 [network graphic 2] (pdf, 26 kB, in German). Network analysis tools like UCINET offer further analysis possibilities (http://www.analytictech.com/ucinet/ucinet.htm).

Statistical analysis methods

Statistical analysis methods can be employed if you have collected quantitative (numerical) data using questionnaires, tests, observations, etc. Statistical approaches serve to represent data e.g. as parameters or graphics (descriptive statistics), as well to test difference hypotheses (inferential statistics).

Scale levels

The answer to the question as to which statistical analysis methods are appropriate for data analysis depends on which scale level your data corresponds to.
More information about scale levels.