By Prof. Guy Doron and Dr. Danny Derby, Relationship OCD Research Unit. Reviewed November 2026.
What the ROCI and PROCSI measure
ROCI: relationship-centered symptoms
The Relationship Obsessive-Compulsive Inventory (ROCI; Doron et al., 2012) is a self-report measure of obsessions and compulsions centered on one’s romantic relationship. The scale includes 12 items loading on three relational dimensions: feelings towards one’s partner (e.g., “I continuously reassess whether I really love my partner”), one’s perception of partner’s feelings (e.g., “I continuously doubt my partner’s love for me”), and one’s appraisal of the “rightness” of the relationship (e.g., “I check and recheck whether my relationship feels right”). Participants rate the extent to which such thoughts and behaviors described their experiences in intimate relationships on a 5-point scale ranging from 0 (not at all) to 4 (very much). All subscales, as well as the total score including all 12 items, have been shown to relate to measures of OCD symptoms, measures of anxiety, depression and stress, and relationship measures (Doron et al., 2012). The internal consistencies of the subscales in our sample (Cronbach’s alphas) ranged from .84 to .89. The internal consistency of the entire scale was .93 and its test-retest (9 weeks) reliability was .69.
Original validation: Doron, G., Derby, D., Szepsenwol, O., and Talmor, D. (2012), Tainted love.
PROCSI: partner-focused symptoms
The Partner-Related Obsessive-Compulsive Symptom Inventory (PROCSI; Doron et al., 2012) is a self-report measure of obsessions (i.e., doubts and preoccupation) and compulsion (i.e., checking) relating to one’s partner’s perceived flaws in six domains: physical appearance, sociability, morality, emotional stability, intelligence and competence. Findings indicated that the PROCSI can be coded as a six-factor measure or a one global-factor measure. Participants rate the extent to which such thoughts and behaviors describe their experiences in intimate relationships on a 5-point scale ranging from 0 (not at all) to 4 (very much). All subscales, as well as the total score including all 24 items, have been shown to relate to measures of OCD symptoms, measures of anxiety, depression and stress, and relationship measures (Doron et al., 2012). The internal consistencies of the subscales in our sample (Cronbach’s alphas) ranged from .83 to .87. The internal consistency of the entire scale was .95 and the test-retest (9 weeks) reliability .77.
Original validation: Doron, G., Derby, D., Szepsenwol, O., and Talmor, D. (2012), Flaws and all.
The group means, standard deviations and scoring instructions from the former separate ROCI and PROCSI norms pages are retained below. These research reference values support assessment; they do not establish a diagnosis or determine a relationship decision.
Research reference points and online feedback
Melli and colleagues (2018) studied the Italian ROCI and PROCSI. A subsequent ROCD app trial by Cerea and colleagues (2020) explicitly describes screening values of ROCI greater than 21.5 and PROCSI greater than 17, citing that work. For integer summed scores, these correspond to 22 or more and 18 or more. These are sums, not item averages.
Our online tool uses 21 or more for ROCI and 17 or more for PROCSI as a prompt to consider assessment. At exactly 21 or 17, its feedback identifies the boundary separately and does not call the score above the published reference. This consultation prompt is not a universal diagnostic cutoff, and the Italian reference points should not automatically be transferred to every language or setting.
Melli, Carraresi, Pinto, Caccico and Micheli (2018), Italian validation, 24(3), 251-269 · Cerea and colleagues (2020), screening methods.
Keep questionnaire totals separate from additional distress and functioning questions. The additional questions in our online tool are not a validated severity scale. A low score does not rule out clinically important distress, and a high score does not diagnose OCD or determine a relationship decision.
Other helpful norms for the ROCI & PROCSI
Group Means, Standard Deviations, in Relationship-Centered (ROCI)
| ROCD1 Participants(n=22) | OCD2 Participants(n=22) | Community3 Participants(n=500) | ||||||
| M | SD | M | SD | M | SD | |||
| ROCI | 2.13 | 0.80 | 1.12 | 0.84 | 0.68 | 0.77 | ||
Group Means, Standard Deviations, in Partner-Focused (PROCSI)
| ROCD1 Participants (n=22) | OCD2 Participants (n=22) | Community3 Participants (n=500) | ||||||
| M | SD | M | SD | M | SD | |||
| PROCSI | 1.30 | 0.70 | 0.81 | 0.67 | 0.56 | 0.61 | ||
Table Notes: 1 Clients presenting with ROCD as main reason of referral; 2 Clients presenting with OCD other than ROCD as main reason of referral;3 Individuals from a large community cohort who completed the questionnaires online.
Rule of thumb: Scoring higher than 2SD above the mean of the community cohort or higher than the mean of ROCD clients (whichever is lower) warrants further clinical attention.
Calculating scores for the ROCI & PROCSI
Download the ROCI and PROCSI scoring guide (English, Word DOCX). Includes the scored item sets and a content-based crosswalk to the original 2012 publications.
ROCI scoring: The ROCI form presents 14 statements. Score the 12 items remaining after excluding items 2 and 8; do not reverse-score the excluded items. Divide their sum by 12 for the item mean (0-4), as used in the original research. The equivalent sum ranges from 0 to 48.
The items for the three ROCI subscales are 1) Love For Partner: 1,7,10,14; “Rightness” of relationship: 3,5,9,12; and Love Of Partner: 4,6,11,13
PROCSI scoring: The PROCSI form presents 28 statements. Score the 24 items remaining after excluding items 1, 4, 16 and 23; do not reverse-score the excluded items. Divide their sum by 24 for the item mean (0-4), as used in the original research. The equivalent sum ranges from 0 to 96.
The items for the PROCSI subscales are: Morality: 6,15,17,20; Sociability: 21,18,5,2; Emotional stability: 24,22,12,7; Competence: 28,27,25,10; Appearance: 26,19,14,9; Intelligence: 3,8,11,13
Translations of the ROCI & PROCSI
The Relationship Obsessive Compulsive Inventory [ROCI]
To download the ROCI in English
To download the ROCI in Spanish
To download the ROCI in Italian (translated by G. Melli, C. Carraresi, E. Righini, 2012)
To download the ROCI in Dutch (translated by G. Anholt, 2013)
To download the ROCI in Persian (translated by A. Atary, 2012)
To download the ROCI in Turkish (Translated and adapted by Inozu, M. & Tirak, E. 2015)
To download the ROCI in Albanian (Translated and adapted by Fatlinda Berisha Tahiri, 2016)
To download the ROCI in Hebrew
How to cite this measure: Doron, G., Derby, D., Szepsenwol. O., & Talmor. D. (2012). Tainted Love: exploring relationship-centered obsessive compulsive symptoms in two non-clinical cohorts. Journal of Obsessive-Compulsive and Related Disorders, 1, 16-24.
The Partner Related Obsessive Compulsive Symptom Inventory (PROCSI)
To download the PROCSI in English
To download the PROCSI in Spanish
To download the PROCSI in Italian (translated by G. Melli, C. Carraresi, E. Righini, 2012)
To download the PROCSI in Persian (translated by A. Atary, 2012)
To download the PROCSI in Turkish (Translated and adapted by Inozu, M. & Tirak, E. 2015)
To download the PROCSI in Hebrew
How to cite this measure: Doron, G., Derby, D., Szepsenwol. O., & Talmor. D. (2012). Flaws and All: Exploring Partner-Focused Obsessive-Compulsive Symptoms. Journal of Obsessive-Compulsive and Related Disorders, 1, 234-243.
Validation studies in other languages
Several research teams have tested the ROCI and PROCSI outside Israel. Not every translation listed above has been validated in a published study.
- Italian: In a community sample of 820 people and in 310 people who reported an ROCD diagnosis, the three-factor structure of the ROCI and the six-factor structure of the PROCSI held, with good internal consistency and construct validity (Melli, Carraresi, et al., 2018).
- Turkish: In 292 adults in a relationship, the factor structures of both questionnaires matched the originals, with good reliability and validity (Trak & İnözü, 2017).
- German: In two samples of 409 and 248 people in relationships, the ROCI’s three-factor structure held and was equivalent for men and women. The authors recommend reporting the three subscale scores rather than a total score (Brauer & Borchardt, 2025).
- Persian (Iran): Researchers developed a version of the ROCI adapted to Iranian culture, keeping the original items and adding new ones. In 341 married students it showed a two-factor structure (Ghomian et al., 2021).
References
Brauer, K., & Borchardt, L. (2025). Validation of the German Relationship-Obsessive Compulsive Inventory: Testing the factorial structure, measurement invariance, and external validity. Journal of Clinical Psychology, 81(11), 1143–1154. https://doi.org/10.1002/jclp.70024
Ghomian, S., Shaeiri, M., & Farahani, H. (2021). Designing the Relationship Obsessive-Compulsive Inventory (ROCI) based on Iranian culture and evaluation of its psychometric properties. Iranian Journal of Psychiatry, 16(4), 418–429. https://doi.org/10.18502/ijps.v16i4.7229
Melli, G., Carraresi, C., Pinto, A., Caccico, L., & Micheli, E. (2018). Valutare il disturbo ossessivo-compulsivo da relazione: Proprietà psicometriche della versione italiana di ROCI e PROCSI [Relationship obsessive-compulsive disorder assessment: The psychometric properties of the Italian version of the ROCI and the PROCSI]. Psicoterapia Cognitiva e Comportamentale, 24(3), 251–269. Article page
Trak, E., & İnözü, M. (2017). A new obsessive-compulsive symptom type: The psychometric properties of the Turkish forms of Relationship Obsessions and Compulsions Inventory and Partner-Related Obsessive-Compulsive Symptoms Inventory. Klinik Psikiyatri Dergisi, 20(3), 171–185. https://doi.org/10.5505/kpd.2017.93063