Ideally every cell in the calculator – which represents every school district in your state – would be filled in (computer-types use the term “populated”) with accurate data. However, in the real world it often is not possible to secure complete data for particular variables. These missing data can pose problems in using the calculator.

At this time, because of limitations in the calculator’s software, it is not possible to leave cells blank as those empty cells will not allow the statistical tools in the calculator to work properly.

You have two options to address the missing data problem.

First option:  if more than 20% of the cases – in this case school districts and students – are missing data on the variable in question, you should consider selecting another variable to use in the sampling process.

Example of when to consider selecting another variable:

If 20% of the school districts, which total 20% or more of the students in your state, are missing data on ELL status, it would then be wise to drop that variable and use another variable for which there is less missing data. If this is done, it is advisable to change the column header to avoid confusion regarding which variables you have used.

Second option: if less than 20% of the cases – i.e., less than 20% of the school districts and students -- are missing data on the variable in question you can address the problem statistically. Stated otherwise, if less than 20% of the school districts are missing the data in question – and those districts comprise less than 20% of the students associated with that variable -- it should be possible to calculate the average value for your state based on those reporting school districts and substitute those values for the missing data.

Example of addressing the missing data problem statistically:

Data are available for 80% of the school districts regarding the number of special education students on IEPs who dropped out of school in the previous year and 20% of the districts did not report the number of dropouts who were special education students. Because the missing districts comprised less than 20% of the high school students in your state, it would then be possible to use the statistical substitution of those missing data.

To calculate the dropout rate for students on IEPs in your state you would divide the total number of special education students on IEPS who dropped out of school in those districts by the total number of special education students on IEPs in the 14 to 21 year old age range in those districts to determine the average dropout rate.

The districts that did report dropout data indicated that 400 special education students on IEPs dropped out of school in the previous year and there was a total of 2000 special education students on IEPs in the 14 to 21 age range in those districts.  The average state dropout rate would then be 400/2000 which would equal 20%. The best bet for those districts that did not report dropout data – the “missing” districts -- would be to apply this rate to those districts.

The calculation you would follow is illustrated here. District A had 50 special education students in the 14 to 21 year old age range. Multiplying 20% by 50 would estimate that 10 special education students in District A dropped out of school in the previous year (.2 X 50 = 10). Obviously, you cannot be sure that 10 students dropped out of District A, but the calculation will be as close as possible to the true figure for that district.