The Life Orientation Test

What it is

The scale is cheap to administer, which is the first fact about it and also the reason the literature is so large. Scheier's own explanation, quoted in Sumit Paul-Choudhury's The Bright Side, is that the brevity of the scale let a great many researchers attach it to work they were already doing. A six-item questionnaire can be bolted onto a cohort study that was designed for something else entirely, and that is exactly what happened across three decades. A cheap instrument attached to everything produces a large literature before it produces a reliable one.

On conventional psychometric grounds the scale performs respectably. Internal consistency in the two biggest cohorts to use it is adequate rather than impressive: alpha of 0.79 in the Nurses' Health Study measurement of 2004, reported by Lewina Lee and colleagues in 2019, and 0.75 in the Women's Health Initiative, reported by Hayami Koga and colleagues in 2022. The scores are also genuinely trait-like over time. In the veterans' cohort in Lee's paper, which used a different instrument, the five-year correlation of the optimism score was 0.87. That stability is the strongest thing anybody can say for treating optimism as a disposition rather than as a mood on the day of the survey.

The problem is not reliability. It is what the summing assumes. Adding a positively framed half to a reverse-scored half treats optimism and pessimism as one quantity with two ends, so that a person low on the scale is simply a person short of optimism. Scheier and colleagues tested that assumption directly in 2020, for the Optimism/Pessimism Meta-Analytic Consortium, and the design is worth describing because it is unusually honest in construction. Subscale numbers cannot be pulled out of already-published papers, so the consortium wrote to the original authors of 61 separate samples and asked them to rerun their own analyses with the two halves separated, then pooled what came back: 221,133 people.

Against an aggregated measure of physical health outcomes, the combined index gave a correlation of 0.026. The absence of pessimism gave 0.029. The presence of optimism gave 0.011, and the difference between the two halves was statistically significant, Z = -2.403, p < .02. The paper's own public significance statement puts it plainly, that the absence of pessimism is more strongly related to good physical health than the presence of optimism is. Their sensitivity analysis then found two studies that, when removed, made the optimism subscale non-significant altogether, while the combined scale and the pessimism subscale depended on no single study. Egger's regression was significant for the combined scale and for the pessimism subscale, so some publication bias was present in the literature, with trim-and-fill leaving the conclusions unchanged. Over 90 per cent of the pooled participants were white and women, and over 90 per cent of the studies were conducted in the United States.

The same asymmetry turns up when the scale is split inside a single cohort. In 159,255 Women's Health Initiative participants, Koga and colleagues found the most pessimistic quartile at 5.2 per cent shorter life span, on an interval from 4.3 to 6.1, against the most optimistic quartile at 2.5 per cent longer, 1.6 to 3.4. The pessimism half is more than twice the size of the optimism half, in their own words slightly larger in magnitude, and in the same direction as the consortium's result.

One further point matters for how the headline finding gets described. The Life Orientation Test is not the only instrument in this literature. The men in Lee and colleagues' paper were scored on the Revised Optimism-Pessimism Scale, derived from 263 items of the Minnesota Multiphasic Personality Inventory-2, with a Kuder-Richardson 20 of 0.88. The two instruments correlate only moderately in prior work, which the paper says. So the much-quoted replication across the two cohorts is a replication across populations rather than across one measure.

In effect

A questionnaire administered once, in 2004, against a ten-year death record

The clearest research case is Lee and colleagues 2019. Optimism was measured a single time, with the revised scale, in 69,744 women in the Nurses' Health Study, and deaths were then tracked to 1 June 2014. The authors excluded 1,400 women who died within the first two years, specifically to reduce the chance that imminent health decline had shaped the questionnaire answers. Adjusted for demographics, baseline conditions and depression, the top quartile showed a 14.9 per cent longer life span, 11.9 to 18.0; with health behaviours added, 8.7 per cent, 5.8 to 11.6. Six items, one administration, and an estimate that nearly halves depending on which model you read.

A self-test printed in a trade book, with scoring bands nobody published

The Bright Side, Canongate, 2025, prints the revised ten-item scale at page 31 as a reader self-test, calling it "the LOT" throughout and never using the term LOT-R. The author reports his own maximum score of 24 and, to his credit, his own doubt about whether he answered honestly or wrote down what he thought he should score. The page also prints bands that sort a reader into low, moderate and high optimism. Those bands have no published provenance, they are not Scheier and Carver's, and they are not reproduced here or anywhere else on this site. The example is worth keeping because it shows what the instrument becomes once it leaves the cohort study: a six-item transparent self-report, easy to guess the intent of, with a category attached to the total.

Employee wellbeing programmes that sell the top of the scale

The commercial use of this instrument and its relatives is in workplace resilience and wellbeing provision, where a short questionnaire measures the trait before and after a course designed to raise it. Two published sentences bound what such a programme can claim. Alan Rozanski and colleagues state in their 2019 meta-analysis that further research will need to assess whether optimism enhanced or induced through intervention carries health benefits similar to optimism that is naturally occurring. And Paul-Choudhury reports, at page 80 of a book arguing for optimism, that optimism-boosting interventions "have mostly been of limited value in boosting optimism among the psychologically healthy". A pre and post score on this scale measures movement on the scale. It is not evidence about anybody's health.

What it does not say

It does not say that a score on this questionnaire is clinically meaningful. There are no published scoring bands, no clinical cut-off, and nothing on this page should be used to interpret anybody's total. The scale was published as a continuous measure.

It does not say that optimism causes better health. The instrument is a measurement, and the designs it has been used in are overwhelmingly observational. There is no Mendelian randomisation study and no co-twin or sibling fixed-effects study of dispositional optimism and mortality, and both absences were searched for before they were stated here.

It does not say that the summed score is worthless. The combined index does predict physical health outcomes, at a correlation of about 0.026 across 221,133 people, and the big cohorts that use it are large and unusually well adjusted. What the evidence will not support is reading a result from the combined score as a result about optimism specifically, when the separated halves put almost all of the predictive work at the pessimistic pole.

And it does not say anything about illness or recovery at the level of an individual person. Nothing here implies that a patient's attitude determines whether they get better. Where this construct has been tested against cancer survival, the positive pole has not predicted it.

One gap in our own sourcing, stated plainly: Scheier, Carver and Bridges 1994, the paper that published the revision, has not been read by this publication. The item wording and the absence of published bands rest on secondary description, and that is why this page is marked mixed rather than robust.

Who developed it

Michael Scheier and Charles Carver published the Life Orientation Test in 1985, and the 1994 revision with Michael Bridges is the version in use. Scheier is also first author of the 2020 consortium reanalysis that separated the scale's two halves and reported the optimistic one as the weaker predictor.

---