Skip to content
Compressed Alignment
Menu

Evidence

What this draws on, and what it does not claim

The assessment is research-informed and provisional. Here is the research behind each area, and the limits.

Direction and prioritiesStructure
Strategic consensus (Kellermanns et al., 2011); goal specification in team processes (Marks, Mathieu & Zaccaro, 2001).
Roles and interfacesStructure
Role ambiguity (Rizzo, House & Lirtzman, 1970); top management team role structure (Ma, Kor & Seidl, 2022).
Decision rightsStructure
No established measure exists for executive decision rights. These questions were written for this assessment.
Cadence and timingStructure
Shared temporal cognition (Gevers, Rutte & van Eerde, 2006); temporal leadership (Mohammed & Nadkarni, 2011).
Information flowRelationships
Information exchange as part of top-team behavioural integration (Simsek et al., 2005).
Challenge and voiceRelationships
Team psychological safety (Edmondson, 1999); task conflict in teams (de Wit, Greer & Jehn, 2012).

The status of the assessment

Each area corresponds to a construct with published measures. The questions in this assessment are not those measures. They were written for executive teams in transition, to be answered quickly and candidly by people who have little time and good reason to be guarded.

The assessment has not been psychometrically validated, and nothing here should be read as saying it has. Its results are indicative: a structured, confidential starting point for a conversation the team needs to have, not a diagnosis.

Why disagreement is measured

For each area the report gives the team’s overall level and how far its members agree. Agreement is measured between people, not between questions, using an index in the tradition of within-group agreement measures (James, Demaree & Wolf, 1984). It compares how spread the team’s answers are with how spread they would be if people answered at random.

Low agreement is reported as a finding. If half a team believes something is settled and half does not, that is often the most useful thing a report can show.

Two teams with the same average and different levels of agreementSix people in each team answer whether decision rights are clear. In team A, three say partly and three say mostly. In team B, three say not clear and three say clear. Both teams average exactly between partly and mostly.Team Amostly agreeNot clearPartlyMostlyClearaverageTeam Bsplit down the middleNot clearPartlyMostlyClearaverage
Both teams average the same. Only the spread shows that Team B has two incompatible views of whether the matter is settled.Illustration, not client data

The hypothesis being tested

The practice works from a hypothesis: after a transition, structural alignment can be rebuilt faster and more deliberately than relational alignment. If that holds, it matters for what a team should do with limited time before a fixed demand. Repeat measurement of the same teams over time is how it is being tested.

Research in progress

The method is being studied in a longitudinal case study of a private-equity-backed organisation whose leadership team went through three CEO periods in about a year. It was presented at ICBML, Oxford, in 2026, as Compressed Alignment: Leadership Team Realignment Under Investor Time Pressure in a PE-Backed Organization. The analysis is still under way. The study treats CREST as a framework to be tested and refined, not an established result, and keeps research interviews separate from coaching.

Publications

Dr. Joanna Konstantinou’s published research and conference papers are listed on her ORCID record. Anonymised assessment data is used for research only where a participant has separately agreed to it.

Questions about the method?

Happy to go through how the assessment was built and what it can and cannot tell you.

Get in touch