Descriptive Statistics

Sample Mean

x̄ = (1/n) ∑i=1n xi

Sample Variance

s2 = (1/(n-1)) ∑i=1n (xi - x̄)2
Uses n-1 (Bessel's correction) for unbiased estimation.

Standard Deviation

s = √s2

Coefficient of Variation

CV = (s / x̄) × 100%

Skewness

g1 = (1/n) ∑ [(xi - x̄)/s]3

Kurtosis (Excess)

g2 = (1/n) ∑ [(xi - x̄)/s]4 - 3

Statistical Inference

Standard Error of Mean

SE = s / √n

Confidence Interval for Mean (t)

x̄ ± tα/2, n-1 × (s / √n)

One-Sample t-Test

t = (x̄ - μ0) / (s / √n)
df = n - 1

Two-Sample t-Test (Pooled)

t = (x̄1 - x̄2) / √[sp2(1/n1 + 1/n2)]
sp2 = [(n1-1)s12 + (n2-1)s22] / (n1 + n2 - 2)

Paired t-Test

t = d̄ / (sd / √n)
d̄ = mean of differences, sd = std dev of differences

Chi-Square Test

χ2 = ∑ (O - E)2 / E

Regression Analysis

Simple Linear Regression Slope

b1 = ∑(xi - x̄)(yi - ȳ) / ∑(xi - x̄)2

Intercept

b0 = ȳ - b1

Coefficient of Determination

R2 = 1 - SSres/SStot = SSreg/SStot

Pearson Correlation

r = ∑(xi - x̄)(yi - ȳ) / √[∑(xi - x̄)2 ∑(yi - ȳ)2]

Control Chart Formulas

X-bar and R Chart

X-bar Chart Limits

UCL = X̄ + A2
CL = X̄
LCL = X̄ - A2

R Chart Limits

UCL = D4
CL = R̄
LCL = D3

Individuals and Moving Range Chart

I-MR Chart Limits

UCLX = X̄ + 2.66 MR̄
LCLX = X̄ - 2.66 MR̄

UCLMR = 3.27 MR̄
LCLMR = 0

p Chart (Proportion)

p Chart Limits

UCL = p̄ + 3√[p̄(1-p̄)/n]
CL = p̄
LCL = p̄ - 3√[p̄(1-p̄)/n]

c Chart (Defects)

c Chart Limits

UCL = c̄ + 3√c̄
CL = c̄
LCL = c̄ - 3√c̄

Control Chart Constants

n A2 A3 D3 D4 d2 B3 B4
21.8802.65903.2671.12803.267
31.0231.95402.5741.69302.568
40.7291.62802.2822.05902.266
50.5771.42702.1142.32602.089
60.4831.28702.0042.5340.0301.970
70.4191.1820.0761.9242.7040.1181.882
80.3731.0990.1361.8642.8470.1851.815
90.3371.0320.1841.8162.9700.2391.761
100.3080.9750.2231.7773.0780.2841.716

Process Capability

Cp (Potential Capability)

Cp = (USL - LSL) / 6σ
Measures spread relative to specification width. Does not account for centering.

Cpk (Actual Capability)

Cpk = min[(USL - μ)/3σ, (μ - LSL)/3σ]
Accounts for process centering. Cpk = Cp(1 - k) where k = |μ - m|/[(USL-LSL)/2]

Pp (Performance Index)

Pp = (USL - LSL) / 6s
Uses overall standard deviation s rather than within-subgroup σ.

Ppk (Performance Index)

Ppk = min[(USL - x̄)/3s, (x̄ - LSL)/3s]

Statistical Tables

t-Distribution Critical Values (Two-Tailed)

df α=0.10 α=0.05 α=0.02 α=0.01
52.0152.5713.3654.032
101.8122.2282.7643.169
151.7532.1312.6022.947
201.7252.0862.5282.845
251.7082.0602.4852.787
301.6972.0422.4572.750
401.6842.0212.4232.704
601.6712.0002.3902.660
1201.6581.9802.3582.617
1.6451.9602.3262.576

Standard Normal (Z) Critical Values

Confidence Level α zα/2
90%0.101.645
95%0.051.960
99%0.012.576
99.9%0.0013.291

Chi-Square Critical Values (Right-Tail)

df α=0.10 α=0.05 α=0.025 α=0.01
12.7063.8415.0246.635
24.6055.9917.3789.210
36.2517.8159.34811.345
47.7799.48811.14313.277
59.23611.07012.83315.086
1015.98718.30720.48323.209
1522.30724.99627.48830.578
2028.41231.41034.17037.566