Research & Evidence

Cat Shyness, Fear and Approach: What Three Surveys Found

By
By Clara LindqvistCertified Canine Behavior Consultant (CCBC) & Feline Environmental Specialist
Read time
3 min read
Published
Conceptual illustration of a cat pausing near a doorway while a visitor sits at a respectful distance; not a study photograph.

Does a shy cat always want to keep away, or can caution and approach show up together? A 2026 study tested that idea through three sequential surveys of cat owners. The clearest result was a strong link between owner-rated shyness and fear, while the main predicted curve between shyness and approach was not statistically supported. Later exploratory analyses identified male cats with high scores on both fear and approach, but that pattern is not a validated personality type.

Conceptual illustration of a cat pausing near a doorway while a visitor sits at a respectful distance; not a study photograph.
A conceptual scene about cautious approach. It does not depict a study cat or establish what an individual cat is feeling.

What the researchers wanted to know

Hattori, Saito and Yamamoto asked whether cat shyness might combine fear with a tendency to approach, as has been proposed for human infants. The authors predicted that approach scores would be higher at both low and high ends of shyness, creating a curved relationship. They tested that prediction using owner questionnaires rather than placing cats in a direct behavior experiment.

The team adapted questions from two instruments designed for human infants and young children: the Colorado Childhood Temperament Inventory and the Infant Behavior Questionnaire-Revised. Owners rated shyness on a 1-to-5 scale and fear and approach on 1-to-7 scales. These scores describe how owners interpreted behaviors they had seen. They do not directly measure a cat's private experience.

Conceptual illustration of a caregiver quietly observing a cat at home; the study used owner questionnaires, not this scene.
The surveys asked owners to rate behaviors they had observed in daily life. This illustration is not a study photograph.

Three surveys, three different samples

The paper reports three sequential surveys, each with its own eligibility frame. Study 1 included 82 cats: 44 male and 38 female, all spayed or neutered and between 12 and 84 months old. Owners were recruited through social media in April 2024; respondents were women aged 20 to 50.

In Study 1, shyness and fear scores moved together strongly (Spearman's rs = 0.788, p < 0.001). The predicted curved relationship between shyness and approach was not statistically significant (quadratic term p = 0.170). A relationship between two ratings does not show that one causes the other, and a non-significant test does not prove that a pattern is impossible.

Study 2 focused on 114 neutered male cats aged 12 to 84 months. The positive association between shyness and fear appeared again (rs = 0.846, p < 0.001). The predicted curve between shyness and approach again was not statistically significant (p = 0.269). An exploratory Ward clustering divided scores into four groups; one included 48 cats with high fear and high approach ratings.

Study 3 surveyed another sample of 114 neutered male cats reported as born stray. It examined score patterns against a duration the paper calls "months of breeding." The article does not define that translated phrase clearly enough to paraphrase it as a precise measure of time living with people. The study did not follow the same cats over time, so it cannot establish how an individual cat changes after adoption.

Conceptual illustration of a cat watching a familiar person with a clear path to approach or retreat; no personality label is implied.
The paper scored fear and approach separately. This illustration does not represent measured data or diagnose an individual cat.

Evidence-led insights

1. One label can hide more than one measured pattern

Reported finding: shyness ratings were strongly associated with fear ratings in the first two surveys, while the planned curved relationship with approach was not confirmed. Interpretation: a word such as "shy" can summarize several observations, but it cannot tell us by itself what a particular cat wants in a specific moment. The study recorded owner ratings, not a direct reading of motivation.

2. An exploratory group is not a confirmed category

The exploratory clustering in the male-cat samples produced a group with high fear and high approach scores. In the stray-born sample, that group included 35 of 114 cats. These are groupings produced by one exploratory analysis using these scales and samples. They do not estimate how common the pattern is among cats generally, and they do not overturn the non-significant primary quadratic tests.

3. The later surveys asked narrower questions

The samples of 82, 114 and 114 had different eligibility rules. The later surveys examined patterns from earlier analyses, but they did not repeatedly measure one group of cats over time. In the stray-born sample, age-adjusted correlations between the reported duration measure and shyness, fear or approach scores were not statistically significant (rs = 0.002, p = 0.986; rs = 0.014, p = 0.886; and rs = 0.150, p = 0.113, respectively). The paper calls that duration "months of breeding," a translated phrase it does not define precisely.

4. Consistent answers are not the same as a validated cat measure

In the first survey, Cronbach's alpha was 0.95 for shyness, 0.88 for fear and 0.81 for approach. Those values describe how consistently the items within each scale aligned in this sample. They do not establish that questions developed for human infants measure the same underlying trait in adult cats. The authors explicitly say that construct validity in cats needs further evaluation.

What the study cannot settle

These were owner reports from self-selected online respondents, not direct behavioral observations by independent researchers. The samples were limited to neutered cats aged 1 to 7 years, and the second and third surveys included male cats only. Respondents were women aged 20 to 50. The article does not clearly state where respondents lived, so the authors' Japanese affiliations do not establish the sample's geographic reach. The adapted questions, owner interpretations and exploratory group analysis all limit how far the results can be generalized.

The study does not provide a test for labeling an individual cat, and it does not evaluate a training, handling or care plan. For readers, the useful distinction is modest: notice the situation and the behavior you can observe, rather than treating one moment as proof of a fixed personality or an internal state.

Source and supporting data

Hattori M, Saito A, Yamamoto S. "Exploration of complex shyness and the relationship between fear and approach behaviors in domestic cats (Felis catus)." PLOS ONE. 2026;21(11):e0360035. Read the peer-reviewed study or its DOI record. The publisher lists an S1 workbook with profiles and behavioral scores for Studies 1 to 3.