Skip to content

Business Statistics and Research Methods Short Notes

One-page UGC NET Commerce revision: averages, dispersion, skewness, correlation, regression, probability, distributions, sampling, tests and report writing.

  • 7 min read
  • Updated
On this page · 12 sections
  1. 1. Central Tendency
  2. 2. Dispersion
  3. 3. Skewness and Kurtosis
  4. 4. Correlation and Regression
  5. 5. Probability
  6. 6. Distributions
  7. 7. Research, Data and Sampling
  8. 8. Sampling Distribution and Estimation
  9. 9. Hypothesis Tests
  10. 10. Report Writing
  11. Last-Minute Traps
  12. Where to Go Next

The whole of Unit 5 on one page. For explanations and worked examples, go back to the chapter. Paper 2 has no negative marking (current NTA pattern; confirm in the latest bulletin), so attempt every question; in this unit most marks are lost to a misremembered formula or the wrong test, not to hard arithmetic.

1. Central Tendency

Full topic

ItemFormula or fact
AM (step deviation)\( A + \frac{\sum fd'}{N}h \)
Combined mean\( (n_1\bar{X}_1 + n_2\bar{X}_2)/(n_1 + n_2) \)
AM properties\( \sum(X - \bar{X}) = 0 \); sum of squared deviations least about the mean
GMnth root of product; growth rates, index numbers; zero or negative values not allowed
HM\( n/\sum(1/X) \); speeds over equal distances, equal outlays
AM ≥ GM ≥ HMTwo values: \( GM^2 = AM \times HM \)
Median (grouped)\( L + \frac{N/2 - cf}{f}h \); sum of absolute deviations least about the median; ogives meet at it
Mode (grouped)\( L + \frac{f_1 - f_0}{2f_1 - f_0 - f_2}h \); from histogram
Empirical relationMode = 3 Median − 2 Mean
Partition values\( Q_1 \): N/4; \( Q_3 \): 3N/4; \( D_k \): kN/10; \( P_k \): kN/100; \( Q_2 = D_5 = P_{50} \)

2. Dispersion

MeasureAbsoluteRelative
RangeL − S(L − S)/(L + S)
QD\( (Q_3 - Q_1)/2 \)\( (Q_3 - Q_1)/(Q_3 + Q_1) \)
MDAverage absolute deviation (least about median)MD/average
SDRoot mean square deviation from mean\( \sigma/\bar{X} \)
CV—\( \sigma/\bar{X} \times 100 \); lower = more consistent
FactDetail
SD and origin/scaleIndependent of origin, not of scale; variance scales by the square
Combined SD\( \sqrt{[n_1(\sigma_1^2 + d_1^2) + n_2(\sigma_2^2 + d_2^2)]/(n_1 + n_2)} \)
First n natural numbers\( \sigma = \sqrt{(n^2 - 1)/12} \)
Normal: QD : MD : SD≈ 2/3 : 4/5 : 1 (0.6745σ, 0.7979σ)
Lorenz curveMax O. Lorenz (1905); graphic inequality; Gini 0 to 1

3. Skewness and Kurtosis

ItemFormula or fact
Positive skewMean > Median > Mode; tail to the right
Negative skewMean < Median < Mode
Karl Pearson(Mean − Mode)/σ, or 3(Mean − Median)/σ, limits ±3
Bowley\( (Q_3 + Q_1 - 2M)/(Q_3 - Q_1) \), limits ±1
Kelly\( (P_{90} + P_{10} - 2P_{50})/(P_{90} - P_{10}) \), ±1
Moments\( \mu_2 = \sigma^2 \); \( \beta_1 = \mu_3^2/\mu_2^3 \); \( \beta_2 = \mu_4/\mu_2^2 \)
Kurtosis\( \beta_2 \) = 3 mesokurtic (normal); > 3 leptokurtic; < 3 platykurtic

4. Correlation and Regression

Full topic

ItemFormula or fact
Pearson’s r\( \text{Cov}/(\sigma_X\sigma_Y) \); −1 to +1; free of origin and scale
\( r^2 \)Coefficient of determination (share of variance explained)
Probable error\( 0.6745(1 - r^2)/\sqrt{n} \); significant if r > 6 PE
Spearman (1904)\( 1 - 6\sum D^2/n(n^2 - 1) \); ties: add \( m(m^2 - 1)/12 \) per group
Concurrent deviation\( \pm\sqrt{\pm(2c - n)/n} \)
Galton (1886)Coined “regression”
\( b_{YX} \) / \( b_{XY} \)\( r\sigma_Y/\sigma_X \) / \( r\sigma_X/\sigma_Y \)
\( r = \pm\sqrt{b_{YX}b_{XY}} \)Product ≤ 1; all three share a sign; AM of b’s ≥ r
b and origin/scaleFree of origin only
LinesIntersect at \( (\bar{X}, \bar{Y}) \); coincide if r = ±1; perpendicular if r = 0
SE of estimate\( \sigma_Y\sqrt{1 - r^2} \)

5. Probability

Full topic

ApproachKey idea
Classical (a priori)Favourable ÷ equally likely total; Laplace
Relative frequency (empirical)Long-run proportion; von Mises
SubjectivePersonal degree of belief
AxiomaticKolmogorov, 1933
RuleFormula
Addition\( P(A \cup B) = P(A) + P(B) - P(A \cap B) \)
Multiplication\( P(A \cap B) = P(A)P(B \mid A) \); independent: \( P(A)P(B) \)
Bayes (1763)\( P(E_i \mid A) = P(E_i)P(A \mid E_i)/\sum P(E_j)P(A \mid E_j) \)
Expectation\( E(X) = \sum xp(x) \); \( V(aX + b) = a^2V(X) \)

Mutually exclusive events with positive probabilities are never independent.

6. Distributions

BinomialPoissonNormal
TypeDiscreteDiscreteContinuous
Formula\( \binom{n}{r}p^rq^{n-r} \)\( e^{-\lambda}\lambda^r/r! \)\( Z = (X - \mu)/\sigma \)
Meannpλμ
Variancenpq < meanλ = meanσ²
ShapeSymmetric only if p = 0.5Positively skewedSymmetric, \( \beta_2 = 3 \)
NameJacob Bernoulli (1713)S.D. Poisson (1837)de Moivre (1733), Laplace, Gauss

Normal areas: ±1σ 68.27% · ±2σ 95.45% · ±3σ 99.73% · ±1.96σ 95% · ±2.58σ 99%. Inflection points at μ ± σ; tails never touch the axis.

7. Research, Data and Sampling

Full topic

ItemKey fact
DesignsExploratory (flexible) · descriptive/diagnostic (rigid) · experimental (causal)
Fisher’s principlesReplication, randomisation, local control
Formal designsCRD · RBD · Latin square · factorial
Primary / secondaryFresh / already collected; check reliability, suitability, adequacy
Questionnaire / scheduleRespondent / enumerator fills
ScalesNominal, ordinal, interval, ratio; Likert, semantic differential, Stapel, Thurstone, Guttman
ClassificationGeographical, chronological, qualitative, quantitative
Sturges\( k = 1 + 3.322\log_{10}N \)
Table partsNumber, title, head note, captions (columns), stubs (rows), body, footnote, source
Probability samplingSimple random, systematic, stratified, cluster, multistage
Non-probabilityConvenience, judgement, quota, snowball
Sampling errorOnly in samples; falls as n rises
Non-sampling errorIn census too

8. Sampling Distribution and Estimation

ItemFormula or fact
SE of mean\( \sigma/\sqrt{n} \); FPC \( \sqrt{(N - n)/(N - 1)} \)
SE of proportion\( \sqrt{PQ/n} \)
SE of difference of means\( \sqrt{\sigma_1^2/n_1 + \sigma_2^2/n_2} \)
CLTSampling distribution of the mean → normal as n grows (n ≥ 30 working rule)
Estimator propertiesUnbiased, consistent, efficient, sufficient (CUES; Fisher)
Unbiased varianceDivisor n − 1
CI for mean\( \bar{x} \pm z\sigma/\sqrt{n} \) (1.645 / 1.96 / 2.58)
Sample size\( (z\sigma/E)^2 \), round up

9. Hypothesis Tests

Full topic

ItemKey fact
Type IReject true H₀; α; producer’s risk
Type IIAccept false H₀; β; consumer’s risk; power = 1 − β
Critical zOne-tail 1.645 (5%), 2.33 (1%); two-tail 1.96, 2.58
DecisionReject H₀ if p ≤ α
TestUsedfOriginator
zLarge n or known σ——
tSmall n, unknown σn − 1; n₁ + n₂ − 2; pairs n − 1Gosset (“Student”), 1908
F / ANOVA3+ means; two variancesk − 1, N − kR.A. Fisher
χ²Counts: independence, goodness of fit, variance(r − 1)(c − 1); k − 1 − paramsKarl Pearson, 1900
UTwo independent groups, ranks— (\( U_1 + U_2 = n_1n_2 \))Mann–Whitney, 1947
H3+ independent groups, ranksχ² with k − 1Kruskal–Wallis, 1952
Rank correlation testSignificance of Spearman’s Rt with n − 2Spearman

Twins: independent t ↔ U · paired t ↔ Wilcoxon signed-rank · one-way ANOVA ↔ H · Pearson’s r ↔ Spearman’s R.

χ² facts: never negative; positively skewed; mean = df, variance = 2 df; expected frequency ≥ 5; Yates’ correction for 2 × 2; counts not percentages.

10. Report Writing

ItemKey fact
TypesTechnical (specialists) vs popular (general readers, managers); oral vs written; interim vs final
Preliminary pagesTitle, transmittal letter or certificate, acknowledgements, contents, lists of tables and figures, executive summary
Main textIntroduction → literature → methodology → results → discussion → findings, conclusions, recommendations
End matterAppendices (questionnaire, raw data) → bibliography → index
StepsLogical analysis → final outline → rough draft → rewriting and polishing → final bibliography → final draft
QualitiesClear, objective, accurate, logical, concise but complete, documented, limitations stated
Footnote abbreviationsibid. · op. cit. · loc. cit. · et al. · cf.

Last-Minute Traps

Where to Go Next

Advertisement
Advertisement