The Perceived Stress Scale: How Researchers Measure Stress

The perceived stress scale (PSS) is a short questionnaire that measures how stressful your life has felt over the last month. It does not count what happened to you. It asks how often you felt upset, out of control, unable to cope or on top of things, and turns the answers into one number. Psychologist Sheldon Cohen and two colleagues published it in 1983 as a research tool for studying how stress relates to health.¹

The number only means something next to other people’s numbers. In a national US telephone poll in 1983, the average adult scored about 13 out of 40, and a 2014 survey in Germany found almost the same average.²³

Scoring the standard 10-item form takes two steps. Flip the answers to the four positively worded items, so the highest answer counts as the lowest, then add all ten.⁴

What the Perceived Stress Scale measures

The Perceived Stress Scale measures appraisal: how unpredictable, uncontrollable and overloaded people judge their own lives to be. Cohen, Tom Kamarck and Robin Mermelstein designed it in 1983 because two people can live through the same events and feel very different amounts of strain.¹

Earlier stress research mostly used life-event checklists, which add up things such as a bereavement or a move. The 1983 paper argued that such lists miss ongoing strain, worries about the future and events nobody thought to list, and that the stress a person feels should predict health better than a tally of events. The authors built on the psychologist Richard Lazarus, for whom an event becomes stressful when you judge it demanding and your resources too thin.¹ For the biology behind that judgment, read what stress is and what it does to the body.

None of the questions mentions work, money or family, so the same form can be given to almost anyone. The standard version has six negatively worded items. They ask how often you were upset by something unexpected, felt unable to control important things, felt nervous and stressed, could not cope with everything you had to do, were angered by things outside your control, or felt difficulties piling up beyond you. Four positively worded items ask how often you felt confident handling personal problems, felt things were going your way, could control irritations, and felt on top of things. Every answer runs from never (0) to very often (4).⁵

A manager who had a calm month at work but a frightening hospital visit with a parent would rate that month on how it felt overall. The form never asks which part of life caused the strain. Finding the source takes a second tool, such as a one-line note of what the month held.

The original, standard and brief versions

The Perceived Stress Scale exists in three lengths: the original 14-item form, a shorter standard version that most research now uses, and a brief version for phone interviews. Cohen and Gail Williamson recommended the standard version in 1988 after testing all three in a national US poll. Four of the original items had weak links to the rest, and dropping them slightly improved internal consistency, the extent to which the items move together.²

A 2012 review of validation studies by Eun-Hyun Lee in South Korea ranked the versions the same way: the standard version performed best and the brief one worst. Lee also found that most validation samples were students or workers and that test-retest reliability was rarely checked. Only the abstract of that review was available to us.⁶

A total of 24 sits above the midpoint of the standard form’s range but below the midpoint of the original’s, so the same number can describe a strained month or a fairly calm one. Before comparing your total with a study or a website, check which version it used.

How is the Perceived Stress Scale scored?

The standard Perceived Stress Scale is scored in two steps: reverse the four positively worded items, then add all ten answers to get a total from 0 to 40. Reversing means a “very often” counts as never and a “fairly often” as almost never, so a high total always means more perceived stress, according to the scoring notes from Cohen’s laboratory.⁴

1. 1Answer all ten items0 = never, 4 = very often, about the last month
2. 2Reverse items 4, 5, 7 and 80 becomes 4, 1 becomes 3, 2 stays 2
3. 3Add the ten numbersa total from 0 to 40

The reversal exists because the positive items point the other way: feeling on top of things very often is a sign of low stress. Someone who felt strained “fairly often” and in control “almost never” has had a hard month, and the arithmetic below records that only when the four positive answers are flipped.

Forgetting the reversal makes a hard month look ordinary. If you use an online calculator, check that it names the 10-item version and reverses the four positive items; on paper, reverse first and add second.

1. Six strain items: how often you were upset, nervous, angered or swamped; each answer from 0 (never) to 4 (very often) counts as given
2. Four coping items: how often you felt confident, on top of things or in control; these are reversed, so 4 counts as 0 and 3 as 1
3. One total: the ten scores are added to a number from 0 to 40; higher means more perceived stress, and the scale has no official cut-offs

What is a normal score on the Perceived Stress Scale?

There is no single normal score on the Perceived Stress Scale, only averages from particular samples. The best known comes from a 1983 Harris telephone poll of US adults, analyzed by Cohen and Williamson, and the scores around its average were widely spread, with a typical spreadstandard deviation: A measure of how widely values spread around their average: small when most values sit close to it, large when they are scattered. Effect sizes are often given in standard deviations so that results measured on different scales can be compared.Full entry in the glossary of about 6 points either side.

The study

Moderate evidence

Who felt most stressed in 1983: a national phone poll of 2,387 US adults

In a 1983 national probability sample interviewed by telephone, the 10-item scale averaged 13.0. Women scored higher than men, scores fell slightly with age and with more education, and they were generally higher in poorer households. The authors recommended the 10-item version for future research and the 4-item version only where time is very short.²

The sample was large and drawn to represent US adults, but it is now more than 40 years old, and later averages depend on how the questions were asked.

Cohen and Denise Janicki-Deverts showed that problem in 2012. They compared the 1983 poll with two online US surveys from 2006 and 2009, run by polling firms. The work was supported by the consumer products division of Johnson & Johnson, which the authors consult for on stress measurement, though the paper says the company played no role in it. Averages were higher online in almost every group, but because the 1983 poll was by telephone, the authors could not tell whether stress had risen or the survey method explained the gap. In all three surveys, women, younger adults and people with less income or education scored higher.⁷

Germany offers a more recent check from outside the US. A representative 2014 survey of people aged 14 and over found almost the same average as the 1983 US poll, with women scoring slightly higher than men.³ Readers elsewhere should look for a validated version and norms in their own language.

Bar chart with 4 bars; the largest is US women, 2009 online survey at 16.1.

• US adults, 1983 telephone poll13
• German adults and teens, 201412.6
• US women, 2009 online survey16.1
• US men, 2009 online survey15.5

Because these patterns recur, a student in their twenties who scores in the high teens is less unusual, within their own group, than a retiree in their late sixties with the same score. Compare yourself with the closest group you can find, and above all with your own earlier scores.

Low, moderate and high bands are a convention

The low, moderate and high bands printed on many versions of the Perceived Stress Scale were not set by its authors. Cohen’s laboratory says in its FAQ that the scale is not a diagnostic instrument, has no cut-offs, and allows only comparisons between people in the same sample.⁸

The bands spread anyway, and not only on websites. A Columbia University trial record on ClinicalTrials.gov uses the same three bands to describe PSS scores.⁹ They are convenient, but a cut-off implies a meaningful line between neighboring scores, and the scale’s authors drew none. Under the common bands a 26 counts as moderate and a 27 as high, yet the two totals can differ by one answer moved a single step, from sometimes to fairly often.

Myth

On the 10-item scale, 0-13 is low stress, 14-26 moderate and 27-40 high, so a high band means you need treatment.

Fact

The scale's authors say it has no cut-offs and is not a diagnostic tool. A score shows how you compare with others who took it and with your own earlier answers.

When a website puts a band next to your score, treat it as a rough label and give more weight to how your score moves from one month to the next.

Is the Perceived Stress Scale valid and reliable?

The Perceived Stress Scale is reliable and valid enough for research, with gaps. In the 1983 validation studies, with two student samples and a community smoking-cessation group, the items hung together well, and scores predicted later health-center visits and physical symptoms better than life-event checklists did. Scores also change, as a measure of the last month should. Answers given two days apart agreed closely, while answers six weeks apart agreed much less, and the authors expected the scale’s predictive power to fade after four to eight weeks.¹ A score from March therefore says little about July, and tracking stress means answering afresh each month.

Validity has a known weak spot: overlap with depression. In 1983 the scale was strongly linked to a depression questionnaire, yet each still predicted physical symptoms on its own, which suggests they measure related but different things.¹ The German study found the same overlap with short depression and anxiety screeners, and two related clusters inside the scale, one about feeling helpless and one about confidence in coping.³ For that reason a high score cannot separate a stressful month from low mood, and only a health professional can tell the two apart.

Why a score cannot diagnose anxiety, depression or burnout

The Perceived Stress Scale cannot diagnose anxiety, depression, burnout or any other condition, and it cannot tell you what is causing your stress. It measures one thing, how stressful the past month has felt.⁵

A high score can come with an anxiety disorder, low mood, grief, a newborn, a hostile workplace or a month of exams. The number looks the same in each case. Burnout has its own questionnaires, such as the Maslach Burnout Inventory, which measures emotional exhaustion, depersonalization and a sense of personal accomplishment.¹⁰

The scale also says nothing about the body. A physical signal such as heart rate variability as a stress marker answers a different question with its own limits. Treat a high PSS score as a reason to ask what lies behind it.

Can I take the Perceived Stress Scale myself?

Yes. Cohen’s laboratory at Carnegie Mellon University posts all three versions on its website, with scoring instructions. Permission for research use is requested through the ePROVIDE platform of the Mapi Research Trust, and the lab says requesting it is free. Translations belong to the researchers who made them.⁵ Use in profit-making ventures, such as corporate trials, needs special permission and a nominal fee, and the lab says it cannot vouch for the accuracy of the translations it shares.⁸

For personal use, the scale can serve as a monthly check-in. Taking it on roughly the same date each month, with a one-line note about what happened, shows whether a change in score follows a change in your life. A rise of several points after a reorganization at work is a prompt to look at workload and support.

When to get help, whatever your score says

A low score does not rule out needing help, and nobody needs a high one before asking for it. The NHS in the UK and the World Health Organization both advise reaching out when stress becomes hard to manage, and the NHS names situations that cannot wait. The three levels below are our own grouping of that advice.

• Now: if you or someone you know needs help immediately, or you have seriously harmed yourself, call your local emergency number straight away or go to an emergency department, with someone else driving; in the UK the NHS says 999 or A&E.¹¹ Do the same if you could act on thoughts of suicide. People in the US can also call or text 988, the Suicide & Crisis Lifeline, a free and confidential line that answers 24 hours a day.¹² Elsewhere, Find A Helpline lists free crisis services by country.¹³
• Soon: for problems that are urgent without being emergencies, the NHS suggests phoning NHS 111 or asking your GP practice for an urgent appointment, and it advises booking a GP visit when stress feels unmanageable or the things you try yourself are not working. The NHS also suggests talking to someone; in the UK, Samaritans answer on 116 123.¹¹ Outside the UK, start with your doctor or local health service.
• Routine: the WHO’s advice for anyone finding stress hard to cope with is to seek help from a trusted health-care provider or another trusted person in their community.¹⁴ If you have been tracking your scores, take them along so a professional can see the pattern.

The bottom line

The Perceived Stress Scale is a well-tested measure of how stressful the last month has felt, and its score means most when set beside similar people or your own earlier answers. Reverse the four positive items before adding, and read any band as a rough label. If your score keeps climbing or a month feels hard to get through, talk to a doctor or someone you trust.