Identifying Students for Gifted and Advanced Academic Services Using Universal Screening and Local Norms

Happy Diverse Children


This information is also available as a sharable PDF document. Please feel free to download this PDF and share it with others. We also discuss the use of this information in this recorded webinar.


When identifying the students who need advanced academic interventions, program coordinators may wish to use the local student population, rather than national norms, as a comparison group.

According to the National Association for Gifted Children (2019), “Students with gifts and talents perform—or have the capability to perform—at higher levels compared to others [emphasis added] of the same age, experience, and environment in one or more domains” (p. 1). When comparing a student’s performance to that of others, defining those “others” as peers in the local student population is a reasonable choice.

This approach is especially useful in schools that serve economically, linguistically, ethnically, and/or racially diverse populations. It is also useful if a district coordinator is concerned about inequitable identification of students who need advanced interventions across campuses.

The following instructions offer two approaches to this process. In the first method, we will assess an entire local population of students using a standardized measure (achievement, aptitude, or another performance measure). Raw student scores will be converted to a percentile rank (a percentage representing how many scores are less than or equal to the student’s score). In the second method, we’ll use the mean and standard deviation of our local population to calculate standard scores (z-scores) and convert those to percentiles.

Percentile Rank for Universal Screening With the Local Population

With universal screening, every student in the local population is assessed. The “local population” can be either all students in a district or, more likely, all students at a school for a given grade level.

You will need a spreadsheet with the raw scores from all students in a grade level. Then, you will calculate the percentage of scores that are less than or equal to each student’s score.

We will use one Excel function to calculate this percentage.

Relative Rank of a Score PERCENTRANK.INC function: “This function can be used to evaluate the relative standing of a value within a data set. For example, you can use PERCENTRANK.INC to evaluate the standing of an aptitude test score among all scores for the test” (Office Support, n.d.-c).

Percent Rank Spreadsheet

Download Spreadsheet: Percent Rank for Universal Screening.xlsx

Local Norms

In the next section, we discuss normal distributions and how we can standardize our local data.

One powerful aspect of local norms is that scores from different performance measures can be converted to standard scores for an apples to apples comparison among measures. Raw scores are meaningless when we are comparing performance on multiple measures. A 42 on one measure may indicate a top performance while an 82 on another measure may indicate low performance. However, by converting raw scores to standard scores, we have useful comparatives.

In the next section, we’ll show you how to do this. We’ll also discuss this process when applied as a common practice in gifted education, “two-stage” identification.

Two-Stage Identification Process

A “two-stage” identification process involves a referral stage followed by a more formal assessment of the referred students. Using this “two-stage” identification process is common in many schools. However, if not properly administered, it has the potential to introduce inequity (McBee et al., 2016). When teachers or others are asked to refer students for advanced academic services and only those students are evaluated, you run the risk that some children who may need advanced academic services are not referred for consideration. Those teachers or other individuals who will be referring students for consideration should be well trained to recognize the advanced academic potential in students from all backgrounds.

When creating local norms, you first need to establish the mean and standard deviation for each grade level of a representative sample of the local population. As noted above, the “local population” can be either students in a district or, more likely, students at a school. Generally speaking, because schools within a district can have different characteristics (e.g., socioeconomic, demographic, etc.), it is common to develop norms for each school within a district.

Once calculated, you can then use the mean and standard deviation of the local population to calculate standard scores and convert those to percentiles for students being considered for advanced academic services.

Selecting Students

Ideally, you would test an entire school’s student population to establish local norms; however, this may not be practical. When it is not feasible to test all students in your local population, you can test a representative sample of that population in order to establish the mean and standard deviation.

Selecting a Representative Population

When selecting students for the representative sample, the goal is to ensure that the students in the sample are representative of the local student population. There is no “rule of thumb” for determining the sample size. You want a large enough sample to ensure you get a normal distribution of scores.

Students can be assigned to the representative sample on a random basis. One approach is to make a list of all students in a grade to be tested and select every Nth student (e.g., every 5th or 10th student) from the list to become part of the local norms. This approach should ensure that the socioeconomic status and demographic characteristics of the sample match the local population as a whole. However, you can verify this by sorting your sample into the relevant groups and verifying that each group is properly represented.

If you are establishing norms for an entire district, you want students at all schools being tested to be included in a proportional manner. For example, if you have 120 fifth-grade students at School A and 60 fifth graders at School B, you would want to use the Nth selection method to choose 24 students from School A’s fifth grade and 12 from School B’s fifth grade.

Mean and Standard Deviation Spreadsheet

First, you need to find the mean and standard deviation for each grade level in your local population or representative sample.

You will create a spreadsheet with the raw scores from all students in a grade level or raw scores from the representative sample of a grade level. For the example below, we consider the raw scores from a standardized measure.

We will use two functions in Excel to determine the mean and standard deviation of the raw scores from the standardized measure.

Mean AVERAGE function: “Returns the average (arithmetic mean) of the arguments. For example, if the range A1:A20 contains numbers, the formula =AVERAGE(A1:A20) returns the average of those numbers” (Office Support, n.d.-a).

Standard Deviation STDEV.P function: We will use this function if we tested all students in the local population. “Calculates standard deviation based on the entire population given as arguments (ignores logical values and text). The standard deviation is a measure of how widely values are dispersed from the average value (the mean)” (Office Support, n.d.-e).

STDEV.S function: We will use this function if we only tested a representative sample of the publication. “Estimates standard deviation based on a sample (ignores logical values and text in the sample). The standard deviation is a measure of how widely values are dispersed from the average value (the mean)” (Office Support, n.d.-f).

Mean and Standard Deviation

Download Spreadsheet: Standard Deviation and Mean.xlsx

Standard Scores and Percentile Ranks

Once you have the mean and standard deviation for your entire local population or a representative sample, you can convert the raw scores for students being considered for advanced academic services into standard scores (i.e., z-scores) and percentile rank.

For this, you will create a spreadsheet with the raw scores from students being considered for advanced academic services.

We will use two functions in Excel to determine the standard score and percentile rank of the raw scores.

Standard Scores (z-scores) STANDARDIZE function: “Returns a normalized value from a distribution characterized by mean and standard_dev” (Office Support, n.d.-d).

Percentile Rank (Cumulative Distribution) NORM.S.DIST function: “Returns the standard normal distribution (has a mean of zero and a standard deviation of one)” (Office Support, n.d.-b).

Standard Scores

Download Spreadsheet: Standard Scores and Percentile.xlsx

Additional Help

We hope this introduction to universal screening and developing local norms will be helpful to you as you identify students to be referred for advanced academic services. Feel free to use or share the spreadsheets in this tutorial with your own data.

Although Prufrock Press does not provide consulting services regarding the topics covered in this tutorial, Laila Y. Sanguras, Ph.D., Baylor University, is available to provide consultation and professional development. Feel free to contact Dr. Sanguras by email.

References

McBee, M. T., Peters, S. J., & Miller, E. M. (2016). The impact of the nomination stage on gifted program identification: A comprehensive psychometric analysis. Gifted Child Quarterly, 60(4), 258–278. https://doi.org/10.1177/0016986216656256

National Association for Gifted Children. (2019). Key considerations in identifying and supporting gifted and talented learners: A report from the 2018 NAGC definition task force.

Office Support. (n.d.-a). AVERAGE function.

Office Support. (n.d.-b). NORM.S.DIST function.

Office Support. (n.d.-c). PERCENTRANK.INC function.

Office Support. (n.d.-d). STANDARDIZE function.

Office Support. (n.d.-e). STDEV.P function.

Office Support. (n.d.-f). STDEV.S function.

Copyright ©2020, Prufrock Press Inc. Permission granted to copy, reproduce, and distribute the content on this web page for educational purposes.