Quant AtlasHomeMarketsAcademyAccount
Quant Atlas · quant-atlas.com
HomeMarketsAcademyAccount
Level II
  1. Academy
  2. Level II
  3. Applied Descriptive Statistics

Level II · Optional · Open anytime

Applied Descriptive Statistics

A compact applied lab for researchers comparing strategy samples. Move beyond a single average to weighting, robust summaries, group composition, conditional rates, and comparable time windows. Every calculation describes the supplied observations; inference and strategy approval require additional evidence.

Study any section in any order. These checks do not affect your core course or assessments. Choice answers and elective progress are tracked in this browser.

Need a refresher? Review Level I descriptive statistics in a new tab.

Optional practice · Browser progress0/4 checks answered
Descriptive statistics elective0% complete
0%

0/2 sections viewed · 0/4 exercises answered

  1. Choose the estimand

Choose a section · All are open
  1. Choose the estimand

Choose the estimand · Section 1 of 2

Weighted means: whose average are you reporting?

A weighted mean is the sum of value times weight divided by the sum of weights. The weights define the question: equal weights describe a typical row, share-count weights describe a typical share, and notional weights describe a typical dollar of exposure. None is universally the correct average.

For descriptive distributions, use nonnegative weights with a positive total. Signed portfolio exposures are useful in other calculations, but do not form an ordinary probability distribution: a negative observation weight can break cumulative-share interpretation. Also distinguish a fee per order from a fee per dollar before aggregating.

A headline average can change because group values change, because group weights change, or both. Decompose these effects before saying that execution or research quality improved. A shift toward smaller, easier orders can make average cost look better without improving the execution of either order type.

Loading the full section…

Per-order and per-share execution cost

One order has 100 shares at 1 bp cost per share; another has 900 shares at 5 bp. Equal-order mean cost is (1+5)/2 = 3 bp. Share-weighted cost is (100×1 + 900×5)/1,000 = 4.6 bp. If prices differ, share weighting is not notional weighting.

  1. Write the target: average per order or per share.
  2. Compute weighted numerator 100 + 4,500 = 4,600.
  3. Divide by total shares 1,000 to obtain 4.6 bp; retain the weighting description.

Weighted mean = Σ(wᵢxᵢ) / Σwᵢ

Keep in mind: Name the unit of observation and the weighting scheme before interpreting an average.

A mistake to avoid

Averaging subgroup means equally generally fails when their observation counts or economic exposures differ.

Link to this sectionSearchGlossary & formulas ↗

Pause and try two quick checks.

You can move on at any time. Each answer explains the reasoning.

Quick check · Weighted mean

Two observations have weights of 100 and 900. What's their effective sample size, and what does it mean?

Quick check · Weighted mean

Trades cost 1 on a notional of 100, and 45 on a notional of 900. What's the mean of the cost ratios, and the ratio of total cost to total notional?

Further reading and documentation
  • NIST/SEMATECH: measures of location ↗
  • NIST/SEMATECH: measures of scale ↗
  • NIST/SEMATECH: exploratory data analysis ↗

You are previewing 2 of 7 sections

The full elective is included with Level II enrollment, together with the core course and the official examination.

Enroll in Level II · $9.99All three levels · $24.99Already enrolled? Log in

Go further · Optional resources

Build your mathematics, statistics, and Python toolkit.

Open these resources at any time. They are separate from the core modules and do not count toward examination eligibility, practice scores, or certification. Programming remains optional.

Ready to explore

Calculus I & II

Study single-variable calculus with worked steps, visual explorations, and immediate-feedback checks: from limits and derivatives to integrals, series, and approximation error, closing with mental math techniques.

Open the optional resource →

Ready to explore

Applied Descriptive Statistics

Compare research samples using weighted summaries, robust location and scale, within- and between-group dispersion, conditional tables, and consistent time windows.

Open the optional resource →

Ready to explore

Intermediate Python for Research

Build reliable research workflows with explained code examples, revealable output, and interactive checks on NumPy, pandas, timing, validation, and reproducibility.

Open the optional resource →