Share this result page with someone you trust to review your answers and result.
{{ profileLead }}
{{ chartNote }}
{{ row.readout }}
| Trait | Mean | Current read | Useful side | Counterweight | Copy |
|---|---|---|---|---|---|
| {{ row.label }} | {{ row.scoreDisplay }} | {{ row.statusLabel }} | {{ row.useText }} | {{ row.counterText }} |
Every row below keeps the original response, keying direction, and keyed score that feeds the three SD3 trait means.
| # | Trait | Prompt | Response | Keying | Score | Signal | Copy |
|---|---|---|---|---|---|---|---|
| {{ row.id }} | {{ row.traitShort }} |
{{ row.shortPrompt }}
{{ row.text }}
|
{{ row.answerLabel }} | {{ row.keyingLabel }} | {{ row.recodedDisplay }} | {{ row.signalLabel }} |
Dark-triad language tries to separate three socially difficult personality patterns that often get collapsed into one harsh label. Machiavellianism describes guarded strategy and willingness to use leverage. Narcissism describes status sensitivity, admiration seeking, and a stronger need to feel important. Psychopathy, in this subclinical research sense, describes lower restraint, faster retaliation, and weaker concern about risk or consequences.
Those distinctions matter because the same outward conflict can have different roots. Someone may plan carefully without being impulsive, seek recognition without being especially retaliatory, or take risks without showing much status concern. A useful SD3 reading starts with the profile shape across the three traits, not with a single verdict about character.
| Trait | What the trait points toward | Common overread |
|---|---|---|
| Machiavellianism | Strategic control, guarded disclosure, leverage awareness, and delayed payback thinking. | Assuming every private or planned action is manipulation. |
| Narcissism | Status sensitivity, recognition seeking, self-importance, and expectations of respect. | Confusing ordinary confidence with entitlement or grandiosity. |
| Psychopathy | Risk tolerance, retaliation, lower restraint, and reduced fear of consequences. | Reading a brief self-report as a clinical or forensic judgment. |
The Short Dark Triad, usually shortened to SD3, is a compact self-report measure for those three subclinical traits. It uses 27 agreement statements, with nine statements assigned to each trait and a 1 to 5 response scale. The result is not one combined dark-triad score. It is a set of three keyed means that are easiest to read side by side.
Self-report personality scores depend on the setting the respondent has in mind. A person thinking about workplace competition may answer differently from the same person thinking about family conflict, close friendships, or a recent argument. Social desirability, fatigue, anger, and impression management can all change which statements feel true in the moment.
Dark-triad terms also carry more stigma than a short scale can justify. SD3 results cannot diagnose narcissistic personality disorder, antisocial personality disorder, psychopathy, or any other condition. They cannot prove intent, predict harm, or settle a dispute about someone else's character. Their safest use is private reflection on the pattern of endorsed statements and the relative separation between the three trait means.
Finish one complete 27-item run before reading the profile. The summary, chart, and item tables appear only after every statement has one response.
Start with the shape of the three means, not the emotional weight of the trait names. A profile led by Machiavellianism with much lower psychopathy describes a different answer pattern from a profile where all three means cluster together. Spread tells you whether the highest trait is clearly separated or only slightly ahead.
The reference labels are context cues rather than normal-versus-abnormal categories. Above reference, Close to reference, and Sharp spread describe the completed response pattern under the displayed reference setup. A high label does not prove harmful intent, and a low label does not prove that the trait is absent.
| Result cue | Reasonable reading | Check before acting on it |
|---|---|---|
| Top trait | The highest keyed mean among Machiavellianism, narcissism, and psychopathy. | Confirm that its strongest item signals fit the same recent setting you had in mind. |
| Lowest trait | The lowest keyed mean, useful as a counterweight when reading the profile. | Look at the numeric gap before treating the rank order as meaningful. |
| Spread | The distance between the highest and lowest trait means; values above 0.60 are labeled sharp. | Treat an even or moderate spread as a mixed profile, even when one trait is first. |
| Largest offset | The trait farthest from the validation-sample reference anchor. | Remember that an offset is descriptive context, not a diagnosis or risk classification. |
| Most endorsed items | The strongest keyed item signals in the completed answer set. | Do not let one provocative item outweigh the nine-item mean and the reverse-scored rows. |
| Lower-scored complements | The lowest keyed item signals, which can temper the top-trait story. | Check whether these rows show genuine restraint, ordinary disagreement, or a mismatch with the setting you had in mind. |
For a like-for-like recheck, keep the instrument, answer style, and reflection setting similar. A comparison across two calm work weeks is easier to read than a comparison between a relaxed weekend and a week shaped by conflict, sleep loss, or performance pressure.
The SD3 is a brief proxy measure for three dark personality traits, not a full personality inventory. Its compact design is part of the point: 27 items, three subscales, nine items per subscale, and a five-point agreement format. That makes the score quick to complete, but it also means item wording and keying direction carry a lot of interpretive weight.
Every trait mean is built from keyed item scores. Direct-scored items keep the selected response value. Reverse-scored items are flipped so that disagreement with an opposite or protective statement raises the relevant dark-triad trait mean. After keying, higher values always point in the same direction: stronger endorsement of the target tendency.
Formula Core
r is the selected 1 to 5 response, k is the keyed score after any reversal, and T is the set of nine items for one SD3 trait.
| Trait | Items | Reverse-scored items | Higher keyed mean usually points toward |
|---|---|---|---|
| Machiavellianism | 1 to 9 | None | Guarded disclosure, strategic manipulation, leverage awareness, and delayed payback thinking. |
| Narcissism | 10 to 18 | 11, 15, 17 | Recognition seeking, status sensitivity, self-importance, and a stronger expectation of respect. |
| Psychopathy | 19 to 27 | 20, 25 | Risk tolerance, retaliation, low restraint, and lower concern about consequences under conflict. |
The displayed reference anchor uses validation-sample means of 3.10 for Machiavellianism, 2.80 for narcissism, and 2.40 for psychopathy. Those values create the dashed chart outline and the largest-offset cue. They do not change the keyed trait means and should not be read as disorder thresholds.
| Output | Boundary | Displayed label |
|---|---|---|
| Profile spread | 0.00 to 0.25 inclusive | Even spread |
| Profile spread | > 0.25 and ≤ 0.60 | Moderate spread |
| Profile spread | > 0.60 | Sharp spread |
| Difference from reference | ≥ +0.75 | Clearer tilt above reference |
| Difference from reference | +0.25 to +0.74 | Above reference |
| Difference from reference | -0.24 to +0.24 | Close to reference |
| Difference from reference | -0.25 to -0.74 | Below reference |
| Difference from reference | ≤ -0.75 | Clearer tilt below reference |
Item-level signal labels use the keyed 1 to 5 value after any reversal: 5 is Very strong endorsement, 4 is Strong endorsement, 3 is Mixed / midpoint, 2 is Lower endorsement, and 1 is Very low endorsement. These labels make the answer ledger easier to scan, but the formal SD3 profile still comes from the three nine-item means.
Psychometric reviews note several limits that matter for SD3 reading. The narcissism subscale leans toward grandiose narcissism, Machiavellianism and psychopathy can overlap in measurement studies, and Likert responses are self-report judgments rather than direct behavioral observations. These limits do not erase the profile signal, but they keep the safest reading descriptive and provisional.
The SD3 describes subclinical tendencies from a completed self-report. It is not a disorder screen, safety assessment, hiring test, forensic measure, or way to label another person. Dark-triad wording can be stigmatizing when it is detached from item evidence and context.
Scoring happens in the browser after the page loads. The completed result can be copied as a link, and the chart or answer ledger can be downloaded or copied. Those outputs may include sensitive answers, keyed scores, and trait labels, so save or share them only when that is intentional.
These examples show how the same three means can lead to different readings depending on spread, keying, and completion status.
A completed run returns Machiavellianism 3.78, narcissism 2.82, and psychopathy 2.31. Top trait is Mach, Lowest trait is Psych, and Spread is 1.47, so the profile has a sharp separation. The useful reading is stronger endorsement of guarded strategy than status sensitivity or retaliatory risk in this run.
Another run returns Machiavellianism 3.25, narcissism 3.00, and psychopathy 3.00. The spread is 0.25, which still reads as Even spread. Machiavellianism is numerically highest, but the lead is too small to treat as a clearly separated trait.
Item 20 says "I avoid dangerous situations" and belongs to the psychopathy set. Selecting Strongly disagree records an original response of 1, but Answer review shows Reverse scored and a keyed score of 5/5. That row is worth checking when the psychopathy mean looks higher than expected.
If the progress label shows 26 / 27 answered, the summary, Trait profile map, trait cards, and exports remain hidden. Use the question navigator to find the statement without a check icon and choose one response.
No. The result is a brief self-report profile of subclinical tendencies. It does not diagnose narcissistic personality disorder, antisocial personality disorder, psychopathy, or any other clinical condition.
Reverse scoring keeps the final direction consistent. For items 11, 15, 17, 20, and 25, the selected response is flipped so higher keyed scores still raise the relevant trait mean.
The map appears only after all 27 statements have an answer. Check the progress label and the question navigator; any row without a check icon still needs one radio response.
Largest offset is the trait farthest from the validation-sample reference anchor. It gives descriptive context for the completed run, but it is not a clinical cutoff.
Yes, if both runs used the same assessment and a similar setting. Compare the three means, spread, and item-level signals before treating a change as meaningful.
Scoring happens in the browser after the page loads. A copied result link can include the completed response pattern, so treat shared links and downloaded files as sensitive.