Interview techniques: a research-backed comparison

The six main interview techniques

1. Structured behavioral interview

The most research-supported approach. Every candidate answers the same predefined questions and is scored against the same rubric. Behavioral questions target past behavior ("tell me about a time...") because past behavior in specific situations predicts future performance better than opinions or hypotheticals. Predictive validity: .51 (Schmidt and Hunter, 1998).

2. Situational interview

Candidates are given a hypothetical work scenario and asked what they would do. Unlike behavioral questions, situational questions do not require prior experience, making them useful for entry-level roles. Predictive validity is slightly lower than behavioral (.45 in most meta-analyses) but substantially higher than unstructured conversations.

3. Competency-based interview

A competency-based interview maps each question to a specific competency the role requires, such as communication, analytical thinking, or stakeholder influence. The rubric scores behavioral evidence of that competency at defined levels. This approach aligns with most HR competency frameworks (SHL, Lominger, Korn Ferry).

4. Work-sample and case interview

Candidates complete a task directly representative of the job: a code review, a case analysis, or a presentation. Predictive validity is the highest of any single method (.54) because the task is the job. The main limitation is cost and time, which AI-assisted workflows can partially reduce.

5. Panel interview

Two or more interviewers conduct the interview simultaneously, reducing individual bias. Panel interviews are more consistent than single-interviewer conversations but still lack the standardization of a defined rubric unless combined with a structured approach.

6. Unstructured interview

The classic conversational interview with no fixed questions and no rubric. Studies consistently show low predictive validity (.38) and high susceptibility to affinity bias, contrast effects, and interviewer fatigue. Unstructured interviews are the most common technique globally, which is the main reason hiring outcomes are inconsistent.

Predictive validity comparison

The validity coefficients from Schmidt and Hunter (1998), a landmark meta-analysis of 85 years of selection research, show:

• Work sample tests: .54
• Structured interview: .51
• Cognitive ability test: .51
• Situational judgment test: .43
• Unstructured interview: .38
• Years of job experience: .18
• Reference checks: .26

These figures represent the correlation between the selection method score and actual job performance. A method with validity .50 explains roughly 25% of the variance in future performance, which is substantial in a noisy human context.

How AI interviewers apply structured techniques

GAIA, Intrvio's AI interviewer, combines structured behavioral and competency-based techniques. The employer defines a question set and a rubric per question, mapping each question to a competency. GAIA conducts the conversation, handles real-time follow-up probes, and produces an evidence-anchored scorecard within seconds of the interview ending. The result is structured interview quality at a fraction of the coordination cost, applied consistently to every candidate in a pipeline.