This article is educational and explains how research works. It is not health advice. Anyone concerned about their own wellbeing may wish to speak with a qualified professional.
Loneliness and being alone are different
Researchers generally separate loneliness from objective aloneness. Being alone describes circumstances: living by yourself, having few contacts, spending a lot of time without company. Loneliness is usually described as a subjective experience, an unwelcome gap between the relationships a person has and the relationships they would like to have. Someone can be alone for much of the week and feel content, while another person surrounded by others can feel lonely.
A related distinction is between solitude, which is often chosen and can be valued, and loneliness, which is typically unwanted. A further term, social isolation, usually refers to a measurable lack of contact or ties. Studies that blur these ideas can produce confusing results, so one useful habit when reading a loneliness claim is to check which of them was actually measured.
Different kinds of loneliness
Some researchers distinguish between emotional loneliness, which relates to missing a close confidant or attachment, and social loneliness, which relates to missing a wider network of friends, colleagues or community. Others describe loneliness as lacking a sense of belonging. These distinctions matter because the same overall score can hide different experiences, and because the ideas of connection discussed in work on social connection have several layers.
How loneliness is measured
Because loneliness is a feeling, it is usually measured by asking people.
Direct single questions. A survey might ask how often someone feels lonely, with answer options ranging from never to often. This is short and easy to include in large surveys, and it allows quick comparisons, but results depend on whether the word “lonely” is used. Some people avoid the term because of stigma, so direct questions may lead to lower reported levels.
Multi-item scales. Established scales ask several questions about feelings such as lacking companionship, feeling left out or feeling isolated from others. Answers are scored and combined. Well-known scales exist in several versions, including short forms made for telephone or general surveys. Scales are generally considered more reliable than a single item, because several questions can average out the quirks of any one, but they are longer and less often included in broad surveys.
Indirect wording. Some instruments avoid the word loneliness altogether and ask about companionship, feeling left out or having someone to talk to. This can reduce stigma effects but measures a slightly different thing.
Researchers check scales for reliability, meaning consistent results, and validity, meaning that the scale measures the concept it claims to. These checks are never perfect, and a scale developed in one country or age group may not work identically in another, particularly when translated.
Why numbers are hard to compare
Headline figures on how many people are lonely vary widely between sources, and the reasons are mostly methodological.
- Different questions and cut-offs. A figure for people who “often” feel lonely will differ from one for those who feel lonely “at least sometimes”. Where a scale is used, researchers must choose a score above which someone is counted as lonely, and that threshold is partly a judgement.
- Different time frames. Asking about feelings today, over the past week or in general can yield different answers.
- Different modes. People may answer differently in a face-to-face interview than in an anonymous online form, in part because of social desirability.
- Different populations. Studies of students, older adults or national samples are not directly comparable.
- Changing context. A single survey taken during an unusual period, such as a time of widespread restrictions on contact, may not reflect typical levels.
For these reasons, trend claims are most credible when the same question, population and method are repeated over time, as in tracking surveys. Comparing one study’s figure against another’s from a different method is often unreliable.
What can and cannot be concluded
Much loneliness research is correlational. It finds that loneliness is associated with other factors, such as living arrangements, health, life events or the quality of relationships. These associations do not by themselves show what causes what. Poorer wellbeing might make people feel lonelier, loneliness might affect wellbeing, or both could stem from shared circumstances. Researchers use longitudinal studies, which follow the same people over time, to try to untangle direction, though these too have limits and cannot fully remove the influence of other factors.
It is also worth being careful with claims about particular age groups being “the loneliest”. Findings differ between studies, partly because of how loneliness is measured and partly because patterns vary across countries and periods. Averages hide a great deal of individual variation, and a group-level pattern says little about any one person.
A short checklist for reading loneliness claims
- Does the study measure loneliness (a feeling), isolation (few contacts) or living alone (a circumstance)?
- Was a validated scale used, or a single question?
- What threshold defines “lonely”, and is it stated?
- Who was surveyed, and how?
- Is the claim about association or about cause?
Applying these questions rarely settles a debate, but it helps in telling a careful finding from a loosely worded headline.