A 35-item fear-of-intimacy scale, and why this is not a list of signs

Published 2025-08-04 · 6 min read · By LovePinnacle Editorial

In 1991, Carol J. Descutner and Mark H. Thelen published a 35-item self-report about anxiety in close dating relationships, and this article will not turn that scale into a list of signs. The paper is "Development and validation of a Fear-of-Intimacy Scale," in Psychological Assessment. A seven-item list would have to quote the questionnaire or invent stand-ins. These notes forbid the items. They also do not contain an average score, so you will not see one here. What you can have is what the validation found, and then a clearly separate research tradition that people often mash together with it.

What the scale was built to ask

The scale is a self-report of anxiety about close dating relationships. Thirty-five items is the length. That is already a reason to distrust any article that offers you seven signs and calls them this instrument. Seven is not thirty-five, and the missing twenty-eight are not a bonus section you can guess. High internal consistency and test-retest reliability are what the authors report. In plain words, the items hung together, and a repeat sitting produced a similar result. "High" is the word they give you. It means the tool behaved like a coherent scale. It does not mean the tool diagnosed anyone.

You might be thinking a coherent scale is a checklist you can run on yourself from memory. You cannot run it from this page, because the items are not here, and they should not be reconstructed. Even if you found the questionnaire elsewhere, a score is not a diagnosis. A licensed therapist can talk with you about anxiety that shows up when closeness increases. This article is education about a validation study. It does not assess you.

What moved when scores moved

Higher scores went with higher scores on a loneliness measure. That is a positive correlation: more anxiety about close dating, more loneliness, in the data they reported. Higher scores went with lower self-disclosure, lower social intimacy, and lower social desirability. Those are negative correlations. The pattern is not "people who fear closeness tell more." It is the opposite direction on disclosure and on social intimacy.

Social desirability matters because some people answer in the way that looks good. The authors partialed that tendency out. The relations with loneliness, self-disclosure, and social intimacy still held. So the links are not explained away as a performance for the researcher. These notes do not give you the size of those correlations. They give you the direction, and they give you the fact that the direction survived that control. Direction without a size is still information. It is not a personal prediction.

Higher scores also went with briefer relationships in self-report, and with therapists' ratings of clients' fear of intimacy. Read those as two different witnesses. One witness is the person filling out the form, describing relationships that did not last as long. The other witness is a therapist rating a client's fear of intimacy. The paper says both lined up with higher scale scores. It does not say the therapists were reading the same 35 items aloud, and it does not say a short relationship proves the fear. Briefer, in self-report, traveled with the score. That is the finding. Your own history may not match the average pattern, and a blog cannot tell you if it does.

A gender-role line these notes barely define

Androgynous subjects had less fear of intimacy than masculine subjects and less than undifferentiated subjects. These notes do not define androgynous, masculine, or undifferentiated. They do not say how those groups were sorted, and they do not give means. Report the comparison and stop. If someone tells you the scale found that "real men fear closeness," they have added a sentence the validation does not contain. The only gender-role result in this packet is the one above: the androgynous group sat lower on the fear score than those two other groups.

A different tradition, not the same test

Hazan and Shaver, in 1987, were not scoring this 35-item scale. Their project was a love quiz in the Rocky Mountain News, and a second study with students. Avoidant was one self-classification in that quiz, not a total on Descutner and Thelen's items. Mixing the two lets you talk as if a newspaper category and a later anxiety scale were one instrument. They are not. Themes can rhyme without the tests being the same.

In the newspaper sample, secure was the largest group, 319 people, 56 percent. The avoidant classification was 145 people, 25 percent. Anxious/ambivalent was 110 people, 19 percent. Twenty-one people skipped the item. Twenty-five checked more than one. Across both of their studies the authors reported approximately 56 percent secure, 24 percent avoidant, and 20 percent anxious/ambivalent. Study 2 was not the newspaper. It was 108 undergraduates, 38 men and 70 women, mean age 18, in a course called Understanding Human Conflict.

Avoidant lovers in that work were described through fear of intimacy, jealousy, and emotional swings from high to low. They were the least accepting. They never had the highest mean on a positive love-experience dimension. On happiness, friendship, trust, and fear of closeness, secure respondents differed from both insecure groups, and the two insecure groups did not differ from each other. Secure readers described the love they wrote about as especially happy, friendly, and trusting, and they emphasized accepting a partner despite faults. Mean reported length was 10.02 years in the secure group, 5.97 years in the avoidant group, and 4.86 years in the anxious/ambivalent group. Average age in all three groups was 36. Divorce in the reports was 6 percent secure, 12 percent avoidant, and 10 percent anxious/ambivalent.

Those lengths are not the "briefer relationships" result from the 1991 scale. They come from a different sample answering a different question. Put them side by side and you will see a resemblance. The resemblance is not permission to treat a Hazan classification as a Fear-of-Intimacy Scale score, or the reverse.

The authors' own limits belong with their numbers. People who sent the quiz back had chosen to take part, and the authors say that group ran a little higher on education and included more women. One classification item packed several issues into a single box. Correlations with parent history were significant but not strong, and likely weaken with age. Child-parent bonds differ from adult romance, because romance is mutual and it usually includes sexual attraction. Some adults revise their models in later relationships. The authors cite Main, Kaplan, and Cassidy, 1985, on parents who broke a cross-generational pattern. They say the results suggest, but do not prove, that the same sorts of forces that affect infant attachment may affect these adult choices. None of that converts their quiz into the 1991 scale.

If closeness makes you anxious, the useful next step is a conversation with a licensed clinician, not a homemade seven-sign test. Signs of emotional intimacy describes a different set of questions about closeness, and it is not a scoring key for either study on this page.

If you have been collecting signs from headlines, will you set the list down and ask a clinician what the anxiety does in your actual relationships?

Two papers, kept apart

• Carol J. Descutner and Mark H. Thelen, "Development and validation of a Fear-of-Intimacy Scale," Psychological Assessment, 3(2), 218–225, 1991. https://doi.org/10.1037/1040-3590.3.2.218
• Cindy Hazan and Phillip Shaver, "Romantic love conceptualized as an attachment process," Journal of Personality and Social Psychology, 52(3), 511–524, 1987. https://doi.org/10.1037/0022-3514.52.3.511

fear · intimacy

← All articles