After you have collected any kind of Observation Codes, INTERACT offers you a range of charts, descriptive statistics, and analysis routines.
| TIP: | Examine the statistic routines, even before you have any real data collected, simply create some sample data with the help of available routines under Start -Files - Open  > Demos, as described in Examine Demo Files. |
Note: For overall analysis, it is best to first merge your separate observation files.
Should you discover that the resulting structure in your merged data files does not provide all the answers you need, INTERACT offers various re-structuring routines to turn you structure up-side-down or sideways, to get the results you need.
IMPORTANT: Be aware that you may need several, differently structured, compilation files with all your data, in order to answer all of your research questions.
But with INTERACT, re-structuring does NOT require to re-do your observations!
Analysis Functionalities
These are the statistics available from mainly from within the Analysis tab:
 | Statistics - This command offers a variation of different statistics. The available routines are described in the section Statistics. |
| Visualisation - The Visualization command offers different visualizations. |
| State-Space-Grid - The State-Space-Grid (Lewis, Lamey & Douglas, 1999) creates a graphical landscape of dyadic interactions, visualizing all overlaps in behavior between all Codes of two different Classes. |
| Contingency Analysis - To create new Events, that cover the timely distances between two Codes; take a look at the Contingency Analysis. |
| Kappa calculation - Compare the reliability of coded Events from at least two observers, with just a few mouse clicks (based on Cohen's Kappa). The INTERACT Kappa routine is explained in Kappa - Inter Rater Reliability.. |
Additional Helpful Routines
Various statistical routines, are not accessible directly from the Analysis tab, partly because they are based on the information as shown in the Timeline Chart and partly because, they are more based on data restructuring than 'real' statistics as such:
| Merge multiple documents - Many analysis routines, become quite time consuming, if you need to perform them, for every single document in your study, separately. Therefore, we advise you to merge all your original documents of one study into one big compilation file. How to do this with the help of INTERACT is explained in Merge Documents. |
| Find co-occurring Codes - To identify all parts, where a specified combination of Codes is coded simultaneously, INTERACT's Co-occurrence Filter is exactly what you need.
To find all Codes that co-occur at some point with a specific behavior, you need the Trim on Code filter. |
| Interval-based exports - To cut your event based data into interval based 'slices', showing you all coded information per predefined time interval, is an often-requested feature. The necessary routines for this can again be found in the Timeline Chart.
Read Interval-based Export for details. |
| Synchronize Data for Comparison - If you want to analyze and compare what happened after a certain Event occurred, take a look at the example described in Synchronize Data for Comparison. |
| Numerical Scale Statistics - If you code scale-based ratings, your Codes can be compared differently than with plain text Codes.
A special routine provides you with additional statistical results, as is described in Rating Scales. |
| TIP: | Remember to consult the corresponding topics for further details! |