Growth Model Probability Curve
Interactive Tool User’s Guide
For Microsoft Excel 2003
October 2009
Contents
Changes from 2007-08 to 2008-09
How do Iopen the file?
How do I use thegraph?
What does thegraph show?
Who is included in the data?
Changes from 2007-08 to 2008-09
The 2008-09 Growth Model Probability Curve includes a district’s students last tested in 2007-08 or 2008-09 and looks back on these students' achievement in earlier years to obtain measures of growth within the district. Some districts may not have sufficient data to support analysis between all possible grade spans. The 2007-08 Growth Model Probability Curve included a district’s students last tested in any year from 2004-05 to 2007-08.
The 2008-09 Growth Model Probability Curve includes enough years of data to support analyses with an initial grade of 3 and a future grade of 8 or an initial grade of 5 and a future grade of 10. However, small districts may not have sufficient data to support these analyses. The 2007-08 Growth Model Probability Curve did not include enough years of data to support these analyses.
How do I open the file?
The file was developed in Excel 2003 and may not function as described below in other versions of Excel.
When you try to open this file, you will see the following error message if the security level for macros in Excelis set to Very High or High. Click OK and close Excel.
Open a blank spreadsheet in Excel, go to the Tools menu, and select Macro - Security.
Then set the security level to Medium and click OK.
Open the file again. You will still see the following security warning. Click on Enable Macros to open the file.
The first time you open the file, you will see the display below. In the future, you will see the display which you saved upon closing the file.
How do I use the graph?
Begin at the Toolbar by selecting a district from the drop down list labeled “Pick a District.” Districts are listed in alphabetical order except for the Average District, which is first in the list.
Next, select a subject (either Math or Reading) from the drop down list labeled “Pick a Subject.”
Next, select the initial grade for your analysis from the drop down list labeled “Pick the initial testing grade.”
Next, select the future grade for your analysis from the drop down list labeled “Pick a grade to see the probability of Meeting.”
The graph will display the data for the selected district as a dark curve. Data for the Average District will display as a light gray curve (see example below). You can position your mouse over either curve to display both the OAKS RIT score and the probability of Meeting for any point.
Below the graph is the score selector, another drop down box that allows you to analyze a specific OAKS RIT score in the initial grade. Select a score from the drop down list to display a second colored vertical bar on the graph highlighting that point.
What does the graph show?
A curve in the graph shows an estimate of the likelihood (“probability”) that a district’s students will Meet Standard in a future grade based on their OAKS RIT scores in aninitial grade. OAKS RIT scores in the initial grade are located on the horizontal x-axis; the probabilities of Meeting in the future grade are located on the vertical y-axis. The OAKS cut score in the initial grade is highlighted by the light blue bar. The probability for Meeting in the future grade at the OAKS cut score in the initial grade is displayed in bold. The RIT score at which students have at least an estimated 50% chance of Meeting in the future grade is located at the point where the red horizontal line crosses the curve.
After you pick a district to analyze, the graph shows two curves. The darker curve displays data for the selected district, while the light gray curve displays data for the state’s Average District. The curve for the selected district may fall to the left of the curve for the Average District,as in the example below. This shows that students in the selected district may be more likely to Meet Standard in a future grade than their peers in the Average District.
You may notice in the above examples that some curves are steeper than others. A steeper curve (such as the first graph above) indicates that the score in the initial grade is a stronger predictor future performance than when a curve rises less quickly (such as the second graph above).
The curve for the selected district may fall to the right of the curve for the Average District, as in the example below. This shows that students in the selected district may be less likely to Meet Standard in a future grade than their peers in the Average District.
If you cannot see the light gray curve, then the selected district may be no different from the Average District, as in this example.
The explanation of the display located above the graph presents the OAKS RIT score at which the selected district’s students have at least an estimated 50% chance of Meeting in the future grade. This point is also marked by the red horizontal line across the graph. The explanation of the display also presents the number of students’ test scores used to generate the curve.
You can select a specific OAKS RIT score for analysis using the score selector, a drop down list below the graph. A bright blue bar highlights the selected score on the graph. The text below the graph presents the selected score and the associated estimated probability for Meeting in the future grade.
Who is included in the data?
The 2008-09 Growth Model Probability Curve is based on test scores for a district’s students last tested in 2007-08 or 2008-09 and looks back on these students' achievement in earlier years. The data includes scores for students in grades 3, 5, 8 and 10 since 2003-04 and for students in grades 4, 6, and 7 since 2004-05.
The probability estimates are based on test scores from a district’s students present in the initial grade on the first school day in May and who remain in the same district until at least the first school day in May of the future grade selected for the analysis. Please note that highly mobile students are not represented in this data. Also note that the probabilities of Meeting three or four years into the future are based on much smaller numbers of test scores than probabilities for Meeting one or two years into the future. Districts typically retain only about 60% of their students over a four year span.
The curve generally represents the last two cohorts of students in the district. For instance, if the initial grade is 3 and the final grade is 6, the data represents all students who were 6th graders in 2007-08 and 2008-09. These students were 3rd graders in 2004-05 and 2005-06, respectively. There are a few grade spans where we have testing histories for only a single cohort of students. For instance, if the initial grade is 3 and the future grade is 8, the data is based on students who were 3rd graders in 2003-04 and 8th graders in 2008-09. Note that this is the first year for which data is available to complete some important analyses, including those with an initial grade of 3 and a future grade of 8 and an initial grade of 5 and a future grade of 10.
A word of caution: Note the footnote at the bottom of the chart (see red box in the screen shot below) which reminds us that this is only one view of student data and should not be taken alone when considering programmatic or evaluative changes in a district.
Oregon Department of Education, Office of Assessment and Information Services, 1 of 8