Can Cohen’s D Be Negative In Psychology? | Read The Sign

Yes, a negative Cohen’s d means the second group scored higher than the first, based on the order used in the formula.

Cohen’s d is a standardized mean difference. It tells you how far apart two group averages are in standard deviation units, not in raw points. A value of -0.60 does not mean the finding is bad, failed, or invalid. It means the difference points in the negative direction because of how the groups were arranged.

The clean formula is:

d = (Mean of Group 1 − Mean of Group 2) ÷ pooled standard deviation

If Group 1 has the lower mean, the numerator becomes negative. If Group 1 has the higher mean, the numerator becomes positive. The size stays tied to the distance between the groups.

When Cohen’s d Is Negative In A Psychology Study

A negative value is common in behavior and social science studies. It appears when the reference group sits below the comparison group. The result may fit the research question perfectly.

Say a stress-reduction group reports lower stress scores than a control group. If the formula is treatment minus control, d will be negative. That can be good news if lower scores mean a better outcome.

• Negative sign: Group 1 scored lower than Group 2.
• Positive sign: Group 1 scored higher than Group 2.
• Zero: The two means are the same, or close enough that the standardized gap rounds to zero.

The APA Dictionary entry for Cohen’s d describes it as an effect-size metric used for power and meta-analysis, with common reference values such as 0.20, 0.50, and 0.80. Those values usually refer to size, not the sign.

Why The Sign Changes

The sign depends on group order. Swap the groups, and the sign flips. The distance stays the same.

If a study reports d = -0.50 for treatment minus control, the same numbers written as control minus treatment would give d = +0.50. The finding has not changed. Only the direction label changed.

What The Absolute Value Tells You

The absolute value strips away the sign. That lets you read the size of the gap. A d of -0.80 has the same size as +0.80. Both sit about eight-tenths of a standard deviation from zero.

The sign answers “which group was higher?” The absolute value answers “how far apart were they?” You need both pieces to read the result cleanly.

The broader APA Dictionary entry for effect size says effect sizes describe the magnitude or meaningfulness of a relationship or result. For Cohen’s d, that magnitude is expressed in standard deviation units between two means.

How To Read Negative Cohen’s d Values

Use the sign, the outcome direction, and the study design together. A negative score on a symptom scale may mean improvement. A negative score on a memory test may mean poorer performance. The number alone is not enough.

Reported d	Plain Reading	Check Before Interpreting
-0.10	Tiny negative difference	Group order may matter more than the size
-0.20	Small negative difference	May still matter for high-stakes outcomes
-0.50	Medium negative difference	Read which group was listed first
-0.80	Large negative difference	Check whether lower scores are better or worse
0.00	No standardized mean gap	Sampling error may still exist
+0.50	Medium positive difference	Same size as -0.50, opposite direction
+0.80	Large positive difference	Read the outcome scale before judging value

Cohen’s old 0.20, 0.50, and 0.80 labels are handy, but they are not laws. A small effect can matter if the outcome has real weight. A large effect can be less useful if the sample is narrow or the measure is weak.

When Negative Is The Desired Result

Negative can be the result a researcher hoped to see. It depends on what the score measures.

• A lower anxiety score after therapy may produce a negative d.
• A lower reaction time can show better task speed.
• A lower error rate can signal better accuracy.
• A lower depression score may point toward symptom reduction.

For outcomes where less is better, a negative d may favor the treatment group. For outcomes where more is better, a negative d may favor the comparison group.

Reporting A Negative Cohen’s d Without Confusion

Good reporting states the group order and the meaning of the scale. Readers should not need to guess why the result is negative.

A clear sentence might read: “The treatment group had lower stress scores than the control group, d = -0.62, because d was calculated as treatment minus control.” That one line gives the direction, the size, and the reason for the sign.

The APA numbers and statistics sheet gives reporting advice for statistical values, including rounding based on precision and reader use. For Cohen’s d, two decimals often work well unless your field asks for a different style.

What To Report	Why It Helps	Clean Wording
Group order	Explains the sign	Calculated as treatment minus control
Outcome direction	Shows whether lower is better	Lower scores mean fewer symptoms
Effect size	Shows standardized distance	d = -0.48
Confidence interval	Shows precision	95% CI [-0.72, -0.24]
Plain sentence	Prevents misreading	The treatment group scored lower

Common Mistakes With The Minus Sign

One mistake is treating a negative value as weaker than a positive one. A d of -0.70 is not weaker than +0.70. They have the same size, with different direction.

Another mistake is calling a negative value “bad.” That label belongs to the research question, not the sign. If lower symptom scores are desired, negative may mean the tested group did better.

A third mistake is dropping the sign in a single study. Meta-analysts may recode signs to align outcomes across papers, but a stand-alone report should preserve direction unless a method note explains the change.

A Simple Check Before You Write It Up

Before reporting the number, ask three plain questions:

1. Which group is first in the formula?
2. Does a higher score mean more, better, worse, or less?
3. Does the sign match the sentence I wrote about the result?

If those answers line up, the negative sign is doing its job. If they clash, the group order, scale direction, or wording needs a fix.

Final Takeaway On Negative Cohen’s d

A negative Cohen’s d is a direction marker. It says the first mean was lower than the second mean under the formula used. The size comes from the absolute value, while the meaning comes from the study design and the outcome scale.

Read the sign, read the group order, then read the scale. That three-part check keeps the result clear and stops a useful finding from being mistaken for an error.