Application Screening Judgment: Reading a CV Without Fooling YourselfApplication Screening Judgment. Most hiring decisions are made at the stage nobody measures.
Two hundred applications become six names, and the whole judgement lives in what each piece of evidence was allowed to count for. This takes that apart, one piece at a time.
Scored as a shape, not as a level
Sixteen exercises put one piece of evidence in front of you at a time and ask a single question: how much should this move you? Three years of closely similar work. A two-year gap with no explanation. A referral from somebody who sat next to them for two years. A highly ranked university for a subject the job never uses. An application sent eleven minutes after the advert went up. A document they produced at work that looks like the work you are hiring for.
Your sixteen answers make a weighting profile, and that profile is compared with what the selection evidence supports by CORRELATION rather than by distance. That difference matters more than it sounds. A distance measure would score a cautious screener who weights everything lightly and a decisive one who weights everything heavily as though their judgement differed, when what differs is their volume. A correlation is invariant to that, so you are judged on which things you weight - and your general level is then reported separately, as its own finding, because over-weighting and under-weighting have different fixes.
Six further exercises test whether one standard survives contact with reality: a late application added as a favour, a shortlist that came out skewed, forty applications in one afternoon, and a file with a fact missing from it. And the report is a receipt. Every claim it makes about you is numbered and answered at the foot of the page by the specific item, your weight and the keyed weight - as endnotes, never as hover text - so you can disagree with any line of it in detail.
What you walk away with
A correlation against the keyed profile, corrected against three thousand simulated screeners drawn from published choice rates on every option.
Whether you weight everything more heavily or more lightly than the evidence supports. A different failure from weighting the wrong things, with a different fix.
Every piece of evidence, your weight, the keyed weight, and the reason - as a table, because sixteen precise values is where a table beats a chart.
The single thing you let count for most and the single thing you let count for least, each with the reasoning behind the key.
Each statement about you footnoted to the thing you actually did that produced it. Endnotes, not tooltips, so the page prints complete.
Inside your report
Illustrative sample - your report is generated from your own responses.
Diamond: what the evidence supports. Dot: what you let it count for.
Built for
- Recruiters and talent teams who screen at volume and want the stage audited
- Hiring managers who make the shortlist without ever having been trained to
- Agencies and RPOs who need a defensible, evidence-based screening standard across a team
Find out what you let count
34 scored exercises - about 40 minutes - a full bespoke report with your weighting match, your elevation reported apart, the full ledger and a numbered receipt for every claim.
₹999 (incl. GST) · assessment and full report, nothing further to pay
Frequently asked questions
Whether the weight you give a piece of application evidence matches the weight that evidence can defensibly carry, and whether one standard survives a late favour, a skewed shortlist and a long afternoon. Four capabilities are reported: what an application can tell you, prestige and similarity and noise, gaps and detours and missing facts, and the same standard twice. It does not measure interviewing, employment law, or whether a particular hire worked out.
By profile-match correlation. Sixteen single pieces of evidence are weighted from decisive to no weight at all, and the vector of your weights is correlated with an authored key. Correlation rather than distance is the point: it makes your score independent of how heavily you weight things in general, and your elevation is then reported separately. The correlation is corrected against three thousand simulated respondents drawn from the declared choice rate on every option, so zero means typical rather than random.
From the published selection literature named in the methodology note - meta-analytic validity of selection methods, halo and similar-to-me effects, employment-gap and tenure research, salary-history anchoring, referral validity, inter-rater agreement, decision fatigue and adverse-impact analysis. The report states in the open that it is an authored key rather than a rated panel with an agreement coefficient, because that panel does not exist yet and implying one would be dishonest.
Different stage and different construct. Interviewer Judgment is about the interview itself - structure, questioning and the biases that operate in a room with a person in it. This one is about the stage before, where a written file is read and a shortlist is produced, and the construct is evidence weighting rather than interview conduct. Teams that hire at volume usually need both, and they measure different people's work.
Rs 999 in India including GST, or US$9.99 elsewhere, one time, for one full sitting and report. Employer assessment suites are annual per-recruiter subscriptions that test the candidate rather than the screener, structured-hiring courses are sold per seat and teach rather than measure, and free bias quizzes test whether you can name a bias. Organisations can use AssessAll credits at 26 credits per person.
Each one takes a single capability, puts you inside the situations where it is actually tested, and scores your choices against published evidence — with a report designed for that capability alone, not a template. They span hiring, compliance, education, operations and personal skill.
Browse the catalogue →Methodology: Measures how a screener weights evidence in a written application, through original situational and selection items keyed to published constructs. Construct statement: it measures whether the weight a person gives a piece of application evidence matches the weight that evidence can defensibly carry, and whether the same standard is applied twice. It does not measure interviewing skill, employment law, recruitment marketing, or whether a particular hire succeeded. Declared response instruction: behavioural tendency throughout - every stem asks what the respondent would actually do or how they would actually weight the item. Item format: an evidence ledger. Sixteen single pieces of evidence are presented one at a time, and the respondent assigns each a weight; the vector of weights is compared with an authored expert vector by correlation, so a respondent who weights everything highly and one who weights everything lightly are not separated by their level but by their pattern. The expert vector is an authored key drawn from the selection literature named below; it is not yet a panel key with a reported agreement coefficient, and the report says so. Sources drawn on across the construct area: meta-analytic work on the predictive validity of selection methods, in which work samples, structured behavioural evidence and job-relevant experience outperform credentials and unstructured impressions; research on the halo effect and its sequential form, where an early impression colours later reading; the similar-to-me effect in selection decisions; prestige and institutional-brand effects on screening; employment-gap research showing that gaps carry little performance information and large penalties; tenure and job-mobility base rates by sector; salary-history anchoring and the evidence behind restrictions on it; referral research on the validity and the access asymmetry of first-hand recommendations; inter-rater agreement in unstructured screening, which is low and improves with written criteria set before reading; decision-fatigue findings across long screening sittings; and adverse-impact analysis, in which a group difference is a prompt to test whether a criterion is job-related rather than a verdict on the process. All items are original works, no commercial screening product or trademarked instrument is named or reproduced, and no affiliation with any source is claimed. AssessAll original design.