Winsteps produces bit-mapped images in the Graph window using the Graphs menu. Winsteps produces character-based graphs in Table 21
Initially, select which Graph window you want to see. When the Graph window displays, you can select the other options. Graphs are plotted relative to the central difficulty of each item or response structure unless "Absolute Scaling" is clicked in the Graphs window . Model-based curves (such as probability and information functions) are the same for all items which share the same model definition in ISGROUPS=. Empirical curves differ across items and rating scales.
Graph window options: |
|
Standard Graph Window |
This is the Graph window implemented in Winsteps since 2002. It is module winvbexe.exe in 32-bit VB6. |
This is the Standard Graph Window with the graph displayed to the bottom right so it can be dragged larger. It is module winvbebr.exe in 32-bit VB6. |
|
Enhanced Graph Window |
This is the Graph window recently implemented in Winsteps. It is module winxojog.exe in 64-bit Xojo. |
Display by scale group |
for each ISGROUPS= code, a set of curves is shown. An example item number is also shown - all other items in the grouping are included in the one set of grouping plots. |
Display by item |
shows curves for individual items. Model-based output is the same for all items with the same ISGROUPS= designation. |
Display by item with DIF ICCs |
Same as "Display by Item", but non-uniform DIF curves are also shown for each item. Enhanced: Uniform DIF curve is also shown |
Vertical histograms showing the the distributions of the person abilities and ... |
|
Person-Item Measure Bar Chart |
item difficulties |
Person-Item Half-Point Thresholds |
rating-scale half-point thresholds between categories: similar to Table 1.5 |
Person-Item Rasch-Thurstonian Thresholds |
rating-scale Rasch-Thurstonian thresholds between categories: similar to Table 1.6 |
Person-Item Rasch-Andrich Thresholds |
rating-scale Andrich thresholds between categories: similar to Table 1.7 |
Person-Item Category Probability Peaks |
rating-scale category centers: similar to Table 1.8, where expected score on the item is the category number, and also the category has the highest probability of being observed. |
Person-Test Measures for Scores |
locations of the measures rescaled to approximate raw scores on the currently active set of items |
