This shows the most unexpected responses sorted by unexpectedness (standardized residual). Large standardized residuals contribute to large outfit mean-square fit statistics. UCOUNT= sets the maximum number of "most unexpected responses" to report in Table 10.6.

 

TABLE 10.6 LIKING FOR SCIENCE (Wright & Masters

INPUT: 75 KIDS, 25 ACTS  REPORTED: 75 KIDS, 25 ACTS, 3 CATS        WINSTEPS 3.58.0

--------------------------------------------------------------------------------

MOST UNEXPECTED RESPONSES

+------------------------------------------------------------------------------------------------------+

| DATA |OBSERVED|EXPECTED|RESIDUAL|ST. RES.|MEASDIFF| ACT   | KID    | ACT          | KID              |

|------+--------+--------+--------+--------+--------+-------+--------+--------------+------------------|

|    0 |      0 |   1.93 |  -1.93 |  -7.66 |   3.53 |    18 |     73 | GO ON PICNIC | SANDBERG, RYNE   |

|    2 |      2 |    .07 |   1.93 |   7.57 |  -3.50 |    23 |     72 | WATCH A RAT  | JACKSON, SOLOMON |

|    2 |      2 |    .07 |   1.93 |   7.57 |  -3.50 |    23 |     29 | WATCH A RAT  | LANDMAN, ALAN    |

|    0 |      0 |   1.93 |  -1.93 |  -7.41 |   3.46 |    19 |     71 | GO TO ZOO    | STOLLER, DAVE    |

 

DATA is the response code in the data file

OBSERVED is the code's value after rescoring

EXPECTED is the predicted observation based on the person and item estimated measures

RESIDUAL is (OBSERVED - EXPECTED), the difference between the observed and expected values

ST. RES. is the standardized residual, the unexpectedness of the residual expressed as a unit normal deviate

MEASDIFF is the difference between the ability and difficulty estimates. This produces the EXPECTED value.

ACT is the item entry number

KID is the person entry number

ACT is the item label

KID is the person label