Research-inspired methods

The Foundations of Our Analysis

How well do we really understand each other? LoveAudit looks at visible timing, language, and interaction patterns to make your conversations easier to reflect on. These signals are not absolute truths, but they can show how your communication feels and where it is worth looking more closely.

01

Chronemics

Edward T. Hall, 1972

Chronemics is the study of how time is used in communication. In digital conversations, response times shape how interest, availability, and prioritization are perceived. Speed and consistency can therefore offer useful signals about rhythm and reciprocity.

How we use this in LoveAudit

We analyze median response times for both participants to determine response-time balance. We use the median instead of the average so individual outliers distort the result less. Gaps over 8 hours are excluded as rest periods. A gap below 20% suggests a similar rhythm.

02

Social Exchange Theory

Thibaut & Kelley, 1959

This theory describes how people perceive attention, time, and energy in relationships. When initiative and effort feel very one-sided over time, that can influence satisfaction.

How we use this in LoveAudit

We look at conversation initiative through three signals: who reconnects after 6+ hours of silence, who sends follow-up messages without a reply, and how text volume is distributed. Together, these signals show how visible effort is shared in the chat.

03

Hyperpersonal Model

Joseph Walther, 1996

Walther discovered that computer-mediated communication often enables more intense relationships than face-to-face encounters, as people can present themselves more selectively. Late-night conversations are particularly significant because social filters and self-presentation concerns tend to drop during these hours, leading to more authentic and emotionally vulnerable exchanges.

How we use this in LoveAudit

We analyze messages between 11 PM and 4 AM as a possible quiet-hours signal. During these hours, daily life is often calmer, which can make personal conversations more likely. For timezone differences, we adjust each person's messages to their local time.

04

Linguistic Style Matching

Ireland et al., 2007

LSM describes the unconscious alignment of language patterns between conversation partners. Studies suggest that similar communication styles can be linked to a stronger perceived connection. This alignment often happens automatically.

How we use this in LoveAudit

We compare two expression categories: laugh expressions (haha, lol, 😂, etc.) and affection markers (❤️, love you, etc.). For each category, we calculate how similarly both people use them. 100% means very similar use; a lower score shows different expression styles.

05

Active Interest & Question Behavior

Charles Derber, 1979

Derber distinguished between shift responses, which move the topic back to oneself, and support responses, which engage with the other person. Questions and follow-ups can show whether interest is actively expressed. A strongly one-sided question ratio can point to unequal participation.

How we use this in LoveAudit

The number of questions asked shows how much active interest each person demonstrates. We count all questions and calculate the ratio. A balanced distribution (close to 50/50) indicates mutual curiosity. We also look at question density to assess overall engagement level.

06

Attachment Theory

Bowlby & Ainsworth, 1969

Attachment theory describes how people manage closeness, security, and distance in relationships. These patterns can influence communication, but they should not be diagnosed from chat data alone.

How we use this in LoveAudit

We detect attachment tendencies through visible chat patterns: closeness-seeking is measured through follow-up messages sent 15 minutes to 2 hours after no reply. Withdrawal is detected when response times exceed 5x the personal median, especially after questions. Consistency is calculated from response-time regularity.

07

Conflict Patterns

John Gottman, 1994

Gottman's research describes stonewalling as a pattern where one person noticeably withdraws from communication during or after conflict. Frequent stonewalling has been associated with strained conflict culture in relationship research.

How we use this in LoveAudit

We detect possible withdrawal patterns by identifying heat phases (8+ messages within 10 minutes with intense content) followed by unusually long silences (4+ hours). To avoid false positives, we exclude nighttime silences, goodnight signals, and configured work hours.

Our analysis shows patterns and tendencies, but it does not replace professional relationship counseling. Algorithms cannot fully capture context, intentions, or the complexity of human relationships.

Ready to understand your patterns?

Upload your chat and see which communication signals are visible in your history.

Start Analysis