Google
 
You are here: Home > SAT Test > ACT Vs. SAT Vs. PSAT > Scoring Methods for Standardized Tests

College Entrance Exams
What Does SAT Stand for?
SAT Scores
SAT Registration
SAT Test Dates
SAT Study Guide
Free SAT Tests
SAT FAQ
What Does ACT Stand for?
ACT Scores
ACT Registration/Test Dates
ACT Preparation
ACT Sample Tests
PSAT/NMSQT
PSAT Scores
PSAT Dates & Registration
PSAT Preparation
PSAT Practice Tests
 

Scoring Methods for Standardized Tests

Different tests use different methods of scoring based on different needs.  The following table summarizes the three main categories of test scores:  

Raw Scores
Criterion-referenced Scores
Norm-referenced Scores (how most standardized tests are scored)

Various Scoring Methods for Standardized Tests: ACT, SAT and PSAT

Score

How Score is Determined

Uses

Potential Drawbacks

Raw Score
 

 By counting the number (or calculating a percentage) of correct responses or points earned.
 		 Often used in teacher-developed assessment instruments.
 		 Scores may be difficult to interpret without knowledge of how performance relates to either a specific criterion or a norm or group.
 

Criterion-referenced Score
 

 By comparing perfoomance to one or more criteria or standards for success.
 		 Useful when determining whether specific instructional objectives have been achieved.  Also useful when determining if basic skills that are prerequisites for other tasks have been learned.
 		 Criteria for assessing mastery of complex skills may be difficult to identify.
 

Age or Grade Equivalent (norm-referenced)
 

 By equating a student's performance to the average performance of students at a particular age or grade level.
 		 Useful when explaining norm-referenced test performance to people unfamiliar with standard scores.
 
  • Scores are frequently misinterpreted, especially by parents.
  • Scores may be inappropriately used as a standard that all students must meet.
  • Scores are often inapplicable when achievement at the secondary level or higher is being assessed.
  • Do not give a typical range of performance for students at that age or grade.
     

Percentile Rank (norm-referenced)
 

 By determining the percentage of students at the same age or grade level who obtained lower scores.
 		 Useful when explaining norm-referenced test performance to people unfamiliar with standard scores.
 		 Scores overestimate differences near the mean and underestimate differences at the extremes.
 

Standard Score (norm-referenced)
 

 By determining how far the performance is from the mean (for the age or grade level) with respect to standard deviation units.
 		 Useful when describing a student's standing within the norm group.
 		 Scores are not easily understood by people without some knowledge of statistics.
 

Table taken from Ormrod, page 530

Standard Scores are by far the most complicated of the five types of scores so they deserve a more in-depth look.  When looking at the normal distribution, a line is drawn from the highest point on the curve to the x-axis.  This point is the mean score.  A standard deviation's worth is counted out on each side of the mean and those points are marked.  Another standard deviation is counted out and two more points are marked.  When the normal distribution is divided up this way, you will always get the same percentage of students scoring in each part.  About 68% will score within one standard deviation of the mean (34% in each direction).   As you move further from the mean, fewer and fewer students will perform at these scores.  A standard score simply tells us where a student scores in relation to this normal distribution in standard deviation units.  There are four common standard scores (Ormrod, 534):

IQ Scores-Tests that measure intelligence have a mean of 100 and (for the most part) a standard deviation of 15.  Most people will score between 85 and 115.  Someone who scores below a 70 is typically considered mentally retarded.
ETS Scores-ETS scores are scores used on tests published by the Educational Testing Service.  They have a mean of 500 and a standard deviation of 100.  No scores fall below 200 or above 800.
Stanines-Stanines is short for Standard Nines.  They are used with standardized achievement tests and have a mean of 5 and a standard deviation of 2.  Each score represents a range of performance (because each score is always reported in whole numbers).
z-scores-These have a mean of 0 and a deviation of 1.  z-scores are typically used by statisticians.

Test Preparation Schools & Programs Letter & Writing Career & Training

Link to us | Contact us | Home Copyright 2008 The Eduers.com. All Rights Reserved