Download presentation
Presentation is loading. Please wait.
1
Using EduStat© Software
7/22/ :10 PM Using EduStat© Software Classic Items Analysis Analyzing the Report © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
2
The report produced following the use of the option "Classic items analysis" includes first of all information on every item (or element of information) under this format: Items Analysis File of data : Number of records: 958 The missing values were included in the calculations Formula corrected for the Point Biserial All records were selected. Multiple-choice items Items Number Prop. Pt.Bis. F F02A F02B F02C F03A .....
3
It is possible to examine the statistical indications calculated for every item (element of information). First, the multiple-choice items are presented. If there is, we find afterward items on partial credits (i.e. on a continuum). 1 Items Analysis File of data : Number of records: 958 The missing values were included in the calculations Formula corrected for the Point Biserial All records were selected. Multiple-choice items Items Number Prop. Pt.Bis. F F02A F02B F02C F03A ..... 1
4
The last two columns of the table presenting the statistics on the multiple-choice items must be considered first and foremost: Prop. This column presents pupils' proportion having answered correctly the question asked; in other words, it is the difficulty of the item. 2 Items Analysis File of data : Number of records: 958 The missing values were included in the calculations Formula corrected for the Point Biserial All records were selected. Multiple-choice items Items Number Prop. Pt.Bis. F F02A F02B F02C F03A ..... 2
5
Pt. Bis. This column reports the values of the biserial point indicating the discrimination of items. The more the value gets closer to "1", the more there is discrimination, that is it is the most successful pupils who answered usually correctly the item and that is the item was failed by the weaker pupils. 3 Items Analysis File of data : Number of records: 958 The missing values were included in the calculations Formula corrected for the Point Biserial All records were selected. Multiple-choice items Items Number Prop. Pt.Bis. F F02A F02B F02C F03A ..... 3
6
The identification of the item
With regard to items on partial credits, the discrimination is indicated by the Pearson correlation of. 4 The statistical report for this type of items includes the following information: The identification of the item 5 The number of records 6 4 Items with continuous answers Items Number Minimum Maximum Average Rés/ Corr. E05A E05B E05C 5 6
7
The minimal value of the mark planned for the item
7 The maximal value of the mark planned for the item 8 The average established with regard to the scale planned (here, it is the value "3") 9 The average result established on 100 10 The value of the coefficient of correlation indicating the discrimination of the item 11 Items with continuous answers Items Number Minimum Maximum Average Rés/ Corr. E05A E05B E05C 7 8 9 10 11
8
Afterward, there is a summary table reporting the discrimination and the difficulty.
12 The last column of the table supplies the distribution of items with regard to the discrimination. So, eleven items are affected by indications of discrimination situated between 0.50 and 0.60 and three between 0.60 and 0.70. 13 12 Summary of discrimination and difficulty index for Extended Response Items Indication of difficulty Discrimination 00>10 10>20 20>30 30>40 40>50 50>60 60>70 70>80 80>90 90>100 Frequencies 0.00> 0.10> 0.20> 0.30> 0.40> 0.50> 0.60> 0.70> 0.80> 0.90> Frequencies 13
9
The last line supplies the distribution of items with regard to their difficulty. So two items have indications of difficulty situated between per cents 10 and 20; five items were made a success by per cents of the pupils. 14 Summary of discrimination and difficulty index for Extended Response Items Indication of difficulty Discrimination 00>10 10>20 20>30 30>40 40>50 50>60 60>70 70>80 80>90 90>100 Frequencies 0.00> 0.10> 0.20> 0.30> 0.40> 0.50> 0.60> 0.70> 0.80> 0.90> Frequencies 14 14
10
The last value registered in the last column of the last line indicates the total number of elements of information. The count of the indications with regard to both examined aspects (difficulty and discrimination) is reported in the various columns and the lines. For example five (5) items have simultaneously an indication of difficulty situated between per cents 50 and 60 of success and an indication of discrimination between 0.40 and 0.50. 15 Summary of discrimination and difficulty index for Extended Response Items Indication of difficulty Discrimination 00>10 10>20 20>30 30>40 40>50 50>60 60>70 70>80 80>90 90>100 Frequencies 0.00> 0.10> 0.20> 0.30> 0.40> 0.50> 0.60> 0.70> 0.80> 0.90> Frequencies 15
11
The last section of the report presents average results for the whole test.
16 Average Standard Alpha Standard err deviation of measure Global Domains Average Standard Alpha Standard err deviation of measure WRITING COPY EXPRESSION SPELLING Skills Average Standard Alpha Standard err deviation of measure APPLICATION SYNTHESIS Common items Average Standard Alpha Standard err deviation of measure COMMON_FRENCH 16
12
In this stage of the data analysis, we should examine first and foremost the results affecting the whole test (Global). Four statistical indications are presented: Average: This indication reports the average obtained to all the items of the test by all the evaluated pupils. The more the average is raised(brought up), the more the pupils made a success of the test. The value of this statistics can be situated between 0 and 100. 17 Average Standard Alpha Standard err deviation of measure Global Domains Average Standard Alpha Standard err deviation of measure WRITING COPY EXPRESSION SPELLING Skills Average Standard Alpha Standard err deviation of measure APPLICATION SYNTHESIS Common items Average Standard Alpha Standard err deviation of measure COMMON_FRENCH 17
13
Standard deviation: The standard deviation is a measure of dispersal of the results obtained by the estimated pupils. The more the value of this indication is low, the more the results obtained by the estimated pupils group together around the overall average. 18 Average Standard Alpha Standard err deviation of measure Global Domains Average Standard Alpha Standard err deviation of measure WRITING COPY EXPRESSION SPELLING Skills Average Standard Alpha Standard err deviation of measure APPLICATION SYNTHESIS Common items Average Standard Alpha Standard err deviation of measure COMMON_FRENCH 18
14
Alpha: This statistical indication indicates the internal coherence of the test. It is a question here of reporting the quality which has a test to measure with the same accuracy to every administration realized in similar conditions, that is the "loyalty". There are several methods to estimate the loyalty of an instrument. The calculation of a coefficient of consistency internal (for example, the coefficient Alpha) is often retained. This coefficient aims at reflecting the degree of association or correlation which exists between variables. 19 Average Standard Alpha Standard err deviation of measure Global Domains Average Standard Alpha Standard err deviation of measure WRITING COPY EXPRESSION SPELLING Skills Average Standard Alpha Standard err deviation of measure APPLICATION SYNTHESIS Common items Average Standard Alpha Standard err deviation of measure COMMON_FRENCH 19
15
Standard error of measure: This indication allows to estimate the margin of error which we have to accept, which error is due to the very mechanism of evaluation, in particular the used instrument. In other words, the standard error of measure is the difference between an observed value and a "true" or real value of what we measure, due to the infidelity of the test or the process of measure. 20 It is possible to realize similar analyses for the groupings of items with regard to the table of specification (domains and skills). Average Standard Alpha Standard err deviation of measure Global Domains Average Standard Alpha Standard err deviation of measure WRITING COPY EXPRESSION SPELLING Skills Average Standard Alpha Standard err deviation of measure APPLICATION SYNTHESIS Common items Average Standard Alpha Standard err deviation of measure COMMON_FRENCH 20
Similar presentations
