Max data to display:

N or n:      ∑x_i:      ∑x²:    RMS=:

Measures/statistics of central tendancy
  middle/center/typical/representative/characteristic/location/summarize:
    Mean x̄ or μ: A sample mean, x̄, is the best point estimate of the population mean, μ. It is unbiased (expected value is μ), consistent (as n↑, x̄→μ), and relatively efficient (smallest variance).

     Median ̃x:     Mode: uni: bi:     Midrange:

    Trimmed mean: 5%(each end):     10%:

    Harmonic mean HM=n/∑(1/x)     Geometric mean GM=ⁿ√Πx

Measures/statistics of dispersion/spread/scatter/variation/variability/uncertainty/volatility /fluctuation/inconsistency:
   Range:
   Standard deviation SD s:    σ:

   Data is "statistically significant" if < x̄-2s= or > x̄+2s=

   Variance VAR s²:    σ²:

   Standard error SEM=s/√n:

   Sample coefficient of variation, CV=s/x̄*100:

   MAD (mean [absolute] deviation): MAD≤σ  Normal:

5-number summary:
Min:    Q1:    Q2(̃x):    Q3:    Max:      IQR=Q3-Q1:
Boxplot:

IQR is range of middle half of data; width of the box.

Outliers:
  Data ≤ Q1-1.5*IQR=
  
  Data ≥ Q3+1.5*IQR=
  

Hildebrand H=(x̄-̃x)/s: If |H|<.2, symmetric. H>.2 right skewed, H<-.2 left skewed
Pearson coefficient of skewness PC=3(x̄-̃x)/s [-3,3]: ~0→symmetric, ≤-1 or ≥1: "significantly skewed"
Skew: |skew|<.5 symmetric, between .5 and 1 moderately skewed, >1 highly skewed.
(Excess) Kurtosis: "tailedness" Normal:0 Uniform:-1.2

Empirical Rule and Chebyshev

% of the data within
1 SD of mean (x̄±s):
2 SDs of mean (x̄±2s):
3 SDs of mean (x̄±3s):

No datum can be more than √(n-1) SDs = from the mean.

Sorted data:

Freq. distr.	Datums	Frequency	Cuml.freq.	Rel. freq.	Cuml.rel.freq

#uniques=

Histogram:     (ogive)

X min: X max: Y max: Class size/width:

Grouped/Classed/Binned frequency distribution:

Class start	Midpoint	Frequency	Cuml.freq.	Rel.freq.	Cuml.Rel.freq.	Norm Expected*n	Normquant Z

                 (Freqs as O_i, Norm Expecteds as E_i into Χ² GOF)

Z scores: σ_n