Arrange= and Juxtapose= control this TableOne table is produced for each arrangement of each facet which lists its elements and their estimates. The arrangement is determined by the Arrange= specification, otherwise it is ascending sequence by element number.
The column headings mean:
Total Score = (when Totalscore=Yes) is the observed raw score (including extreme elements).
Obsvd Score = Observed Score: (when Totalscore=No) observed raw score, the sum of the active responses for that element (after removal of extreme elements and renumbering of rating scales or partial credit scales to eliminate unobserved intermediate categories, when Keep not specified, so that Ordinal is active.)
Total Count = (when Totalscore=Yes) is the observed respond count (including extreme elements).
Obsvd Count = Observed Count: (when Totalscore=No) number of active responses observed for that element
Obsvd Average = Observed Average: Obsvd Score divided by Obsvd Count
Fair Avrge = Fair Average: Rasch measure to raw score conversion, standardized so that it is fair to all elements in the facet. Fair-M uses the facet means as the baseline. Fair-Z uses the facet local origins (zero points) as the baseline.
Measure = Rasch measure/calibration of ability/difficulty/leniency etc. in log-odds units (logits) or user-scaled units based on Umean=
If preceded by an "A", this value was anchored (preset).
If preceded by a "G", this value was part of a group anchor.
For extreme scores, the estimate and standard error are based on the Xtreme= specification. Measures for extreme scores are omitted from the summary statistics. The type of extreme score is also shown:
"Maximum" means an extreme positive estimate.
"Minimum" means an extreme negative estimate.
"Unmeasurable" means no estimate could be obtained, usually because all the observations for this element are part of extreme scores for other elements.
S.E. = Standard Error (precision) of the measure estimate.
Model S.E. = the asymptotic standard error when the data fit the model
Real S.E. = the model error enlarged by data-to-model misfit (infit)
"Standard error=Real" for the larger, data-sensitive standard error
For extreme scores, the standard error is for the estimate obtained using the Xtreme= specification. Standard errors for extreme scores are omitted from the summary statistics.
For paired comparisons, Models=?,-?,.., the standard error is relative to the latent variable, for relative standard errors between elements of the same facet, the standard error is Error/√2.
Infit MnSq = Infit Mean-Square: the information-weighted mean-square fit statistic, with expectation 1, and range 0 to infinity.
Less than 1 indicates muting: too little variation, lack of independence.
More than 1 indicates noise: unmodelled excess variation.
Infit Zstd = Infit Z-standardized: the MnSq statistic standardized toward a unit-normal distribution so effectively a t-statistic with infinite degrees of freedom, i.e., a z-score. The accuracy of this standardization is data dependent. This tests the statistical hypothesis: "Does the apparent randomness in these data fit the Rasch model exactly?" Since no empirical data ever does, results must be interpreted with this in mind. the reported value is trancated towards zero. Thus, any value in the range 1.00 to 1.99 is reported as 1. In the range -1.00 to -1.99 is reported as -1.
Outfit MnSq and Zstd = Outfit Mean-Square and Z-standardized: the outfit statistic has the same form as infit, but is the conventional mean-square which is more sensitive to outliers.
Estim. Discrim = Estimated Discrimination: an estimate of the item discrimination computed according to the "two-parameter logistic model" (2-PL) and "Generalized Partial Credit Model" approach, but without allowing the discrimination estimate to alter other estimates. 1.0 is the expected value. Values higher than 1.0 indicate a steeper than expected empirical ICC, less than 1.0 a flatter empirical ICC. Negative values indicate reverse discrimination. Inspect these on the Graphs menu. According to www.rasch.org/rmt/rmt142a.htm discrimations in the range 0.5 to 1.5 provide reasonable fit to the Rasch model.
PtMea = Point-measure correlation: is produced by "Pt-biserial=Measure". It is the correlations between the observations and the measures modeled to generate them.
PtExp = Expected values of the Point-measure correlation: is produced by "Pt-biserial=Measure". It is the expected correlation between the observations and the measures modeled to generate them, when the data fit the Rasch model.
PtBis = Point-Biserial correlation: is produced by "Pt-biserial=Yes" and is a many-facet version of the point-biserial correlation between responses and total score (including extreme scores). Negative values may imply data problems, but this is highly data-design-dependent.
Exact Agreement = this is produced for the facet specified by Interrater=, See Agreement Statistics.
Obs % = Observed % of exact agreements between raters on ratings under identical conditions.
Exp % = Expected % of exact agreements between raters on ratings under identical conditions, based on Rasch measures.
Displ. = Displacement measure: when a column with this heading is produced, it contains rough estimates of the displacement of the reported measures from those expected if:
i) the analysis had reached convergence, or
ii) the reported measure were not fixed at an anchored value.
Displacements less than the convergence criteria or the measure standard errors are not reported here. They can be seen in the Scorefile=. When there are no displacements to report, this column is not produced.
Displacement = (observed raw score - expected raw score based on reported measure) / (model-derived raw-score variance)
N = number of the element in the specifications
Name of facet is above the element labels.
At the foot of the table is the count of elements, and the means and sample standard deviations for each column. Measures and standard errors for extreme scores are omitted from the summary statistics.
Count is the count of all elements in the facet listed in the Table
Mean is the average of all values in that column listed in the Table
S.D.
To Table 7 reliability and chi-square statistics