PRACTICE GUIDE CORNERSTONE RESEARCH
Cornerstone Research Causal Inference & Competition Practice Test
The Cornerstone Research causal inference & competition screen is 22 questions in 35 minutes - about 1.6 minutes each - answered by typing the number, with no calculator. There is no negative marking, so leaving an item blank gains you nothing over guessing.
Where it sits: First online round. The practice sitting on this page runs the same item count, the same clock and the same marking rule, with questions generated by Outcry rather than taken from Cornerstone Research.
Analyst interviews are widely reported to centre on econometrics fundamentals - regressions, causal inference, and explaining results clearly.
Outcry is not affiliated with Cornerstone Research and has no access to their assessment content. This guide describes an assessment format that candidates report publicly; the questions here are generated by Outcry and are not Cornerstone Research’s own.
What it screens
An economic and financial consulting firm supporting expert testimony in litigation and regulatory proceedings.
- ✓Spotting the confounder that breaks a causal claim
- ✓Difference-in-differences: what the parallel-trends assumption buys you
- ✓Market definition and the hypothetical-monopolist test
- ✓Explaining a regression result to someone hostile to it
Where it sits at Cornerstone Research
This is the whole sitting rather than one section of it: 22 items in 35 minutes, on one clock.
There is no going back. You answer or skip, and the next item loads, which means the decision to leave an item is final at the moment you make it.
The format
These are the numbers the Cornerstone Research sitting on this site runs on, matching the format candidates report.
| Questions | 22 |
|---|---|
| Time | 35 minutes |
| Per question | 1.6 minutes |
| Negative marking | No |
| Answer style | Typed numeric answer, no calculator |
| Where it sits | First online round |
What it tests, with a worked example
Every example below is generated by Outcry, drawn from the same question generators the timed drills run. None of them is Cornerstone Research’s.
Causal inference
Whether the effect you are shown is the effect that happened.
Example
What must an instrumental variable satisfy?
- Correlated with the outcome and the regressor
- Normally distributed with mean zero
- Correlated with the endogenous regressor, and uncorrelated with the error term
- Uncorrelated with the regressor and correlated with the error
Answer Correlated with the endogenous regressor, and uncorrelated with the error term
Relevance plus the exclusion restriction. The instrument may affect the outcome only through the endogenous variable.
Difference-in-differences
The parallel-trends assumption, and what breaks it.
Example
A 10% price rise on product A costs A 29,000 units. Product B gains 11,000 units and product C gains 5,000. What is the diversion ratio from A to B, in percent?
Answer 37.9
D(A→B) = units B gains / units A loses = 11,000/29,000 = 37.9%. The denominator is A's own loss, not the total gained by B and C: 16,000 of the 29,000 units went to B and C and the rest left the market entirely. The same ratio can be built from elasticities, D(A→B) = (ε_BA/ε_AA)·(Q_B/Q_A), which is what gets used when there is no price experiment to read off.
Market definition
Substitution, cross-price elasticity, and where a market's edge actually is.
Damages and counterfactuals
The but-for world, and how much of the difference the conduct explains.
Reading a regression
Coefficients, standard errors and what a control is doing to both.
Example
The maximum likelihood estimate of a normal variance divides the squared deviations by n. With n = 8, how does it compare to the true variance on average?
- Too small by a factor of 0.875
- Too large by a factor of 1.143
- Too small by a factor of 0.125
- Unbiased - maximum likelihood estimators always are
Answer Too small by a factor of 0.875
E[MLE] = ((n − 1)/n)σ² = (7/8)σ² = 0.875σ². Maximum likelihood is consistent, not unbiased, and the gap only closes as n grows: at n = 8 the estimate is 12.5% low on average.
Instrumental variables
Relevance and exclusion, and which of the two an interviewer will push on.
Example 1
Before After Treated (A) 484 377 Control (B) 498 455
What is the difference-in-differences estimate?
Answer -64
(377 − 484) − (455 − 498) = -107 − (-43) = -64.
Example 2
What must an instrumental variable satisfy?
- Correlated with the outcome and the regressor
- Normally distributed with mean zero
- Correlated with the endogenous regressor, and uncorrelated with the error term
- Uncorrelated with the regressor and correlated with the error
Answer Correlated with the endogenous regressor, and uncorrelated with the error term
Relevance plus the exclusion restriction. The instrument may affect the outcome only through the endogenous variable.
Example 3
Demand has constant own-price elasticity of −3 and the firm sets the monopoly price on constant marginal cost. A new input levy raises marginal cost by $4 per unit. By how much does the profit-maximising price rise, in dollars?
Answer 6
Constant-elasticity monopoly pricing gives P = MC × ε/(ε−1) = MC × 1.5. Price is proportional to marginal cost, so ΔP = 4 × 1.5 = $6. Pass-through is 150% of the cost rise, above 100%, because the markup is proportional to cost and rises with it. Assuming full but not over-full pass-through would understate the price effect by $2.
What a good score looks like
On a paper of 22 questions with no penalty for a wrong answer, the only thing an unanswered question can do is cost you. These screens are usually one gate among several and are rarely reported with a threshold. What candidates do report is that the written rationale carries as much weight as the selected option.
How to train for it
- 01Learn the identification strategies by name and by failure mode: what each one assumes, and the one thing that breaks it.
- 02Write a two-sentence rationale for every answer. That is the deliverable, not the letter you picked.
- 03Name the counterfactual out loud on every question. What would have happened otherwise is the question underneath most of these, however they are phrased.
TRAIN IT HERE
The drills that match each section
Causal Confounder
Work out what drove the effect in a diff-in-diff setup with a confounder in it.
Antitrust
Market definition the way a consulting case asks it.
Twenty Backtests
Multiple-testing intuition: why the twentieth backtest always looks great.
SIT THE FULL BATTERY
All the sections back to back on one clock, marked the way the real screen marks them, with a by-skill breakdown at the end. Included with any pass.
Also reported at Cornerstone Research
Common questions
- Is the Cornerstone Research causal inference & competition test multiple choice?
- Typed numeric answer, no calculator. You type the number, so there is nothing to eliminate your way to.
- How long is the Cornerstone Research causal inference & competition test?
- 22 questions in 35 minutes, which is about 1.6 minutes each.
- Is there negative marking on the Cornerstone Research causal inference & competition test?
- No. A wrong answer costs nothing beyond the mark you would have earned, so leaving an item blank is never better than guessing at it.
- How do I practise for it free?
- Every drill linked on this page is free to play, with no account, inside a daily run cap. Questions are generated fresh each run, so there is nothing to memorise between attempts. The full Economic Consulting Analyst Screen sitting puts the sections back to back on one clock.