HSPT Scores Explained: Percentiles, Composites and the Cognitive Skills Quotient

The HSPT (High School Placement Test) is a 298-question exam that high schools, many of them Catholic, give to eighth graders applying for ninth grade. Its publisher, Scholastic Testing Service (STS), reports each of the five subtests as a standard score from 200 to 800 and a national percentile from 1 to 99, adds three composite scores, and also reports a Cognitive Skills Quotient (CSQ), which STS says "replaces the traditional IQ score". STS sets no pass mark: each school decides what a score means. This guide explains what each number is, where the CSQ comes from, what STS says about its comparison group, retests and test preparation, and why an HSPT percentile is not an IQ percentile.

What is on the HSPT?

STS's Interpretive Manual (2021 edition) calls the HSPT "a norm-referenced, standardized test designed for high schools to administer to 8th graders entering high school". It has five multiple-choice subtests that were "designed to be taken in its entirety":

• Verbal Skills: 60 items (synonyms, antonyms, verbal analogies, verbal classifications and logical reasoning)
• Quantitative Skills: 52 items (number series, number manipulations and quantitative comparisons)
• Reading: 62 items
• Mathematics: 64 items
• Language: 60 items (punctuation, capitalization, spelling, grammar, usage and composition)
• Total: 298 items

The first two are the "cognitive skills" subtests, which "require the use of reasoning skills". The other three are "basic skills" subtests, designed "to assess learned skills". Calculators are not allowed on the Quantitative Skills or Mathematics subtests, and there is no penalty for guessing, so every question is worth answering. A school may add one 40-item optional test, Science or Catholic Religion, which is not part of any composite. STS's own pages give no times for the subtests. The Archdiocese of Washington, which runs a diocesan testing program, lists 16, 30, 25, 45 and 25 minutes for the five subtests (141 minutes of testing) and about 3 hours for the whole session, including two short breaks and 30 minutes for handing out and collecting materials. STS's overview page says the HSPT has provided measures "since 1955", while the 2021 manual says it has been "in continuous use since 1958"; STS does not explain the difference.

How are HSPT scores calculated?

Each subtest starts as a raw score, the number of correct answers. STS converts it to a standard score on a scale from 200 to 800, "with a mean of 500 and standard deviation of 100", and to a national percentile, which a sample STS report describes as "the percentage of students in the national sample who had scores lower than yours". The three composites take the same route, but they are not averages. In STS's words, "the raw scores are added together from the subtests that comprise each composite score and are then converted to standard scores and national percentiles". They are Total Cognitive Skills (Verbal plus Quantitative), Total Basic Skills (Reading, Mathematics and Language) and the battery Composite (all five). STS recommends the Composite: research has "demonstrated repeatedly that the battery Composite score is the best single measure for predicting subsequent academic performance".

A standard score of 500 is average on that scale and each 100 points is one standard deviation, a measure of spread explained in our guide to IQ z-scores. If the scores were spread as a normal curve, 600 would sit near the 84th percentile and 700 near the 98th. STS reports the percentile from its own norm sample, though, so read the percentile printed on the report rather than computing one.

What do HSPT percentiles, stanines and grade equivalents mean?

Percentiles run from 1 to 99 and compare a student with others in the same grade. STS calls 50 "exactly average" and treats the 33rd to 66th percentiles as the average band, the middle third of the national sample. The top 10% are rated "High" and the bottom 10% "Low". Stanines, which run from 1 to 9 with a mean of 5 and a standard deviation of 2, give the same information in coarser bands. STS's table puts stanine 9 at percentiles 96 to 99, stanine 7 at 76 to 87 and stanine 5 at 41 to 59.

The comparison group is the catch. The manual says the norm base "is intended to be a more selective base than general population norms". It is built from students "entering high school who would subsequently graduate with a diploma", which STS says "could be called a high school graduate norm base". STS says it conducts norming studies every year, and that the HSPT is normed on eighth graders only: "Other grade level norms are not available". A national percentile of 50 therefore marks the middle of a graduate-bound group, not the middle of all students. A local percentile is narrower again: it ranks a student only against the others a school sent in for scoring.

Grade equivalents appear for Reading, Language and Mathematics only. STS warns against reading a grade equivalent as a placement: an eighth grader with a Reading grade equivalent of 10.7 "was not tested on 10th grade material", and the figure shows only that the student did better than the average eighth grader in the sample.

What is the Cognitive Skills Quotient (CSQ)?

The CSQ is the HSPT's IQ-style number. The manual says: "This measure replaces the traditional IQ score, but its purpose within the school setting remains the same: to function as a predictive index of a student's future academic performance in order to assess learning potential." It is based on the Verbal and Quantitative subtests "as well as his or her age at the time of testing", which is why the answer sheet has to carry a birth date. STS says the scale "has an operational range of 55 to 145 with a mean of 100 and a standard deviation of 15", the convention most IQ tests use. Our guide to standard deviation 15 versus 16 shows why that detail decides whether two tests mean the same thing by the same number.

STS's own guide to reading a CSQ is below. The percentages are STS's wording for the "school population":

• Above 130: approximately the upper 3%
• 110 and above: the upper 25%
• 100 to 109: the 50th to 75th percentiles
• 90 to 99: the 25th to 49th percentiles
• 89 and below: the lower 25%
• Below 70: approximately the lower 3%

Run those cut points through a normal curve with a mean of 100 and a standard deviation of 15 and most agree. A CSQ of 110 is the 74.8th percentile and 90 is the 25.3rd, so STS's quartile boundaries match. A CSQ of 130 is the 97.7th percentile, so 2.3% of scores fall above it, and 70 is the 2.3rd percentile. STS's "approximately 3%" at both ends is either a rounding up or a feature of its own norm group, and its pages do not say which. At the extremes of the scale, 145 is the 99.87th percentile and 55 the 0.13th. Our percentile calculator converts any other value.

Many families never see the CSQ. STS lists it as a standard feature of its school-side reports, the student score labels and the alphabetical lists, but the sample one-page Student Score Report for parents shows standard scores, percentiles and a performance graph, and describes the HSPT as "a measure of basic skills and educational achievement". Which reports a family receives is the school's choice, according to STS's FAQ.

Is the HSPT an IQ test?

No. STS describes the CSQ as a measure of "learning potential" and adds that "unlike pure intelligence tests, however, these subtests do not restrict themselves to measure only innate abilities". The items were designed to measure "the cognitive skills (i.e., skills related to learning) whether such skills are innate or acquired". That is a fair description of most school ability tests, and our guide to achievement tests and IQ tests explains why the line between the two is blurrier than the names suggest.

The CSQ also rests on far less than a clinical IQ test. It comes from 112 multiple-choice items (60 verbal and 52 quantitative) taken in a group, whereas an individually given battery such as the Wechsler scales uses a psychologist and many task types (see WAIS index scores explained). STS says reliability, validity and correlations with other standardized tests are in its annual HSPT Technical Report, which we could not find online, so we cannot say how closely a CSQ tracks an IQ score. Treat it as a school score on an IQ-shaped scale.

What is a good HSPT score?

STS gives no passing score. Its parent page says: "Each school evaluates your scores in its own way. Criteria determined by one school to assist with academic decisions may not be the same for another school." The manual adds that most schools use two or more subtest scores for course placement, with the Quantitative and Mathematics scores considered for math courses and the Verbal, Reading and Language scores for English, alongside "elementary school grades and teacher recommendations". Some reports also carry a predicted ACT (pACT) and an estimated SAT (eSAT) score, "only reported for grade 8 students who take all of the subtests".

For orientation, STS's labels are the safest guide: a national percentile of 50 is average, the middle third is "Average", and the top 10% is "High". A high percentile says a student is ahead of most of a graduate-bound group. It does not say the student has an IQ above 100 in the general population, and the Composite is the figure STS recommends for admissions and scholarships. Our guide to gifted cutoff scores shows how schools often set such thresholds without any single national rule.

Can you retake the HSPT, and does test prep help?

STS "recommends that the HSPT should only be taken once by each student". It adds that "under extenuating circumstances a school may decide to allow a student to retest", case by case. Dioceses can be stricter than STS. The Archdiocese of Washington says a student may not take the exam more than once and that "only the first score will be considered by admissions committees", and the Archdiocese of Baltimore says the test "cannot be repeated under any circumstances". Some online guides say STS uses the lower score if a student sits twice; we found no such rule on STS's pages or on those two diocesan pages, so ask your school directly. On preparation, STS says it "is in no way affiliated with" commercial test-prep products and "believes that parents and teachers are best advised to avoid expensive preparation materials". Its own tips are to pace yourself, read each question, answer every question and check your work. How much practice moves an ability-type score is a separate question, covered in our guide to the practice effect on cognitive ability tests.

How does the HSPT compare with the ISEE and the UK tests?

The ISEE reports four section scores of 760 to 940 with no total and ranks students against recent private-school applicants (see ISEE scores explained). The HSPT ranks students against a national sample built from students expected to graduate, and it reports composites. In the UK, the ISEB Common Pre-Test and CAT4 report standard age scores, which sit on a scale closer to the CSQ's.

To see how an IQ-scale result compares with the population, take the IQ Metrics test, or try the kids IQ test for a child, then read the result against the IQ bell curve.

Common questions