Three questionnaires account for most of what people search for when they look for a "test". They measure related but genuinely different things, and knowing which is which makes your own result far easier to interpret.
By Sam Roberts, MA Professional Counselling; Clinical Supervisor (ACA) — last reviewed 2026-09-05
The PHQ-9 measures the severity of depressive symptoms over two weeks. The GAD-7 measures the severity of generalised anxiety symptoms over two weeks. The DASS-21 measures three separate dimensions at once — depression, anxiety and stress — over one week, and is designed to distinguish between them rather than to screen for a specific condition. All three grade intensity. None of them makes a diagnosis.
The PHQ-9 asks how often, over the last two weeks, nine specific experiences have been present: reduced interest, low mood, sleep change, energy, appetite, self-critical thinking, concentration, physical slowing or restlessness, and thoughts of self-harm. Each item is answered on a four-point frequency scale and totalled.
Its strength is that it maps directly onto the recognised description of a depressive episode, which makes it useful shorthand between professionals and useful for tracking change over time. Its weakness is that several of the items are sensitive to things other than mood — sleep, appetite and energy all move for medical and situational reasons — so a moderate score in someone with a newborn or a thyroid problem means something different from the same score in an otherwise settled life.
A final item asks about thoughts of self-harm. Any endorsement of that item is treated as a reason to speak to someone qualified promptly, regardless of the total.
The GAD-7 asks about seven experiences over two weeks: nervousness, uncontrollable worry, worrying about many different things, difficulty relaxing, restlessness, irritability and a sense that something awful might happen. It was developed to identify generalised anxiety, and it performs reasonably well as a broad indicator of anxiety severity.
Because its focus is generalised worry, it is less sensitive to anxiety that is tied to a specific trigger — social situations, health, panic episodes or past trauma. Someone with intense but narrowly focused anxiety can produce a modest GAD-7 score while still finding daily life significantly restricted, which is a good illustration of why the impact question matters more than the number.
The DASS-21 takes a different approach. Rather than screening for one condition, it produces three separate subscale scores over the past week: depression, anxiety and stress. The stress subscale is the distinctive part; it captures tension, difficulty relaxing, irritability and over-reactivity, which the other two instruments largely leave out.
That makes it useful when someone is clearly struggling but the shape of the struggle is unclear, and particularly useful when the honest answer turns out to be sustained tension and depletion rather than either depression or anxiety. It is a research and clinical instrument rather than a primary-care screening tool, and its scores are read against its own severity bands rather than the ones used by the other two.
If what has changed is mood, interest and energy, the depression-oriented check is the closest match. If what has changed is worry and the inability to settle, the anxiety-oriented check is closer. If everything feels heavy and you cannot tell which of the three is driving it, the three-dimensional approach is the one that will separate them for you.
It is entirely reasonable to complete two. Depression and anxiety co-occur often, and a pair of results gives a fuller picture than either alone. What is not useful is completing five in one sitting, because tiredness starts to shape the answers.
Pay attention to the reference window as well. Two of these instruments ask about the past two weeks and one asks about the past week, which means a single hard week affects them differently. If the last seven days were unrepresentative, note that alongside your result rather than treating the number as settled.
One more practical point: answer for how things have actually been, not for how they are at the moment of answering. People completing a questionnaire on a good evening consistently under-report, and people completing one immediately after an argument consistently over-report. The instructions ask about frequency across the window precisely to smooth that out.
A score tells you how intense a defined set of experiences has been in a defined window, relative to a large reference sample. It does not tell you why. It cannot distinguish grief from depression, or a thyroid problem from low energy, or three months of five-hour nights from a concentration difficulty.
That "why" is the work of an assessment: history, context, the alternatives ruled in or out, and a clinician who has done it many times. A questionnaire arrives at that conversation as evidence, not as a conclusion.
It is also worth remembering that these instruments were designed and normed largely outside Singapore, and that the way distress is described here does not always match the wording of the items. Many people report physical symptoms, tiredness or difficulty at work long before they use the language of mood or worry, and a questionnaire built around emotional wording can under-count that entirely.
If a score sits in a moderate or higher range, or if the difficulty has persisted for more than a few weeks and is affecting work, sleep or your relationships, that is a reasonable point at which to speak to someone.
This article is general information, not clinical advice, and nothing in it diagnoses anything. If what you read here matches your experience closely, that is a reason to speak to someone qualified rather than a conclusion in itself.