Statistics for MBA/ Business statistics explained by example

Examine central tendency, including mean, median, and mode, with arithmetic concepts, plus Excel calculations; and explore dispersion, such as range, variance, standard deviation, coefficient of variation, and z scores.

Learn how to compute the arithmetic mean, the most used measure of central tendency, using ungrouped and grouped data, with step-by-step Excel examples and the average function.

Highlight the advantages and disadvantages of the arithmetic mean as a central measure. It captures information from all observations yet is sensitive to extreme values and open-ended classes.

Learn geometric mean to measure multiplicative growth, shown via a venture capital reinvestment example, calculated as the nth root of the product and applicable to inflation and compound interest.

Calculate the weighted mean by multiplying each labour type's hours by its rate and summing the costs. Divide by the total hours to obtain the average cost per hour.

Identify the median as the middle value in a sorted list. For odd observations, it's the (n+1)/2 item; for even, it's the average of the n/2 and n/2+1 items.

Learn how the median resists extreme values and outliers, its limitations as a central value, and how sorting and comparing with mean help detect outliers.

Define mode as the most frequent value and illustrate its use with grouped data, advantages, and disadvantages; showcase real-world applications like vehicle design and cafe seating.

Explore dispersion and why it matters, showing how data spread affects the reliability of the mean through two datasets and introducing range, average absolute deviation, variance, and standard deviation.

The range equals maximum minus minimum, offering an easy measure of spread, but it ignores variance among data points and is highly sensitive to outliers, making it an imperfect representation.

Learn how to compute the average absolute difference from the mean using absolute deviations, and why weighting extremes via squaring leads to variance as a dispersion measure.

Learn to compute variance and standard deviation from data: subtract the mean, square deviations, sum, divide by n or n-1, then take the square root, with Excel stdev.s examples.

The lecture explains that the sum of squared distances from observations is minimized at the mean, and a constant prediction based on the mean minimizes squared error across all observations.

Compute the coefficient of variation by dividing the standard deviation by the mean to compare variability across units, and use z statistics to standardize observations.

Explore two practical exercises to calculate central tendencies and dispersion using Excel, including mean, median, and arithmetic mean from data and ranges, with an answer sheet for verification.

Explore normal distribution and the central limit theorem, their detailed properties, and how data take their shape, using Excel simulations to verify concepts.

Learn how frequency distribution converts decimals into integer categories (0–99) using floor, counts each category, and uses a chart to reveal a uniform distribution.

Understand the normal distribution as a bell-shaped, symmetric curve where mean, median, and mode coincide. Standardize to z-scores to reveal 68.27% within one sd, 95.45% within two, 99.73% within three.

Normal distribution appears in nature as a bell-shaped curve, symmetric around the mean with most data near the center and fewer at the extremes.

Aggregation drives data toward a normal distribution, demonstrated by summing dice and simulating observations to reveal the central limit theorem for the mean and sum.

Revisit the normal distribution’s bell shape and symmetry, verify the 68.27/95.45/99.73 ranges within one, two, and three standard deviations, and connect to sample means, central limit theorem, and confidence intervals.

Explore sampling distributions by examining means and proportions, review the central limit theorem, and learn point estimates, confidence intervals, and t-distribution concepts for small samples.

Understand sampling terms and their roles, including population, population parameter, sample, sampling statistic, estimator, and test statistic, and learn how random samples estimate population values with the central limit theorem.

Explore sampling statistics through examples of sample mean, median, variance, and standard deviation, and learn how these estimates relate to population parameters and percentiles.

the lecture shows the average of sample means equals the population mean and the standard deviation of sample means is sigma divided by sqrt(n), illustrating the central limit theorem.

Explore how sampling distributions of proportions behave like a normal curve as you take larger samples, using coin toss examples, and learn the standard deviation formula sqrt(p(1-p)/n) for proportion estimates.

Explains the sampling distributions of means and proportions for iid data, showing that the mean of sample means equals the population mean while standard deviations follow sd/sqrt(n) and sqrt(p(1-p)/n).

Learn to use estimation to infer population parameters from sample data, employing point estimates and interval estimates for means and proportions under the central limit theorem.

See how sample proportions follow a normal distribution and lead to a confidence interval. Using two standard deviations around the sample proportion, estimate the population proportion with 95 percent.

Explore confidence intervals with formal definitions and a quick table for large samples, using x-bar and sample proportion to estimate standard deviations and form interval estimates for means and proportions.

Compute a 95% confidence interval for the sample proportion, p-hat = 0.55, from 200 students, as an estimate of the population parameter, using the standard error and 1.96 multiplier.

The lecture explains constructing a confidence interval for the mean using the standard error sigma-hat over sqrt(n), rooted in the central limit theorem, with an Excel-based demonstration.

Demonstrates the central limit theorem with a 200,000 random-number Excel simulation and shows how to form confidence intervals for the mean using two- and three-standard-deviation rules around the population parameter.

Compute the standard deviation of the sample mean by dividing the sample standard deviation by the square root of n, then form the confidence interval using z-values.

Explore how small sample sizes (n<30) rely on the t-distribution rather than the normal, and learn to compare with large-sample CLT, using Excel to illustrate confidence intervals.

Demonstrates the t-distribution with an Excel-based simulation, compares it to the normal distribution, and teaches degrees of freedom and reading the t-table for confidence intervals.

Calculate a confidence interval for a small sample using a t-table with 9 degrees of freedom; for mean 11400 and sd 700, n=10, the 95% interval is 10899 to 11901.

Explore the criteria of a good estimator—unbiased, efficient, and consistent—using the sample mean to estimate population parameters and the impact of sample size on standard deviation.

Explore hypothesis testing with an example, defining null and alternate hypotheses, and cover Type I and II errors, power, P-values, one and two tailed tests, mean and proportion, Excel analysis.

Explore a business hypothesis testing example using copper wire diameter data to illustrate mean, standard deviation, normal distribution, and probability distributions, with Excel demonstrations.

Explain how to use a sample mean and a 99 percent threshold to verify a 4 mm wire claim, and determine when to reject an entire container.

Define null and alternate hypotheses and use sample data to verify population parameters through hypothesis testing, highlighting significance level, critical values, critical and acceptance regions with a diameter example.

Explore the steps of hypothesis testing with a 100-sample example, including null and alternative hypotheses, the sample mean as test statistic, the central limit theorem, and a 0.01 significance level.

Define Type I and II errors and the power of a test, then illustrate with a supplier risk example and central limit theorem using sample means.

Explore Type I and Type II errors, their trade-off, and how increasing sample size boosts test power using the normdist approach and predefined alpha levels.

Explore real life examples of Type I and Type II errors, including false positives and false negatives, with null hypotheses, alpha and beta concepts in diagnosis, court judgments, and tests.

Explore when to apply one-tailed versus two-tailed tests, and how null and alternative hypotheses with one-sided or two-sided critical values determine rejection regions in practical cases like medicine and cables.

Learn how to compute p-values for one-tailed and two-tailed tests from a cricket ball example, including Excel calculations and using the null hypothesis for hypothesis testing.

Demonstrates large sample and small sample hypothesis tests for mean and proportion using football diameter and coin bias examples, with an Excel workbook and step-by-step guidance.

Learn to select the correct statistic for hypothesis tests using a practical table, including z-tests, t-tests, one-sample proportion tests, and a medicine example.

Learn to perform a two-sample t-test in Excel using the data analysis tool pack, including an F-test for variances to decide equal versus unequal variance, and interpret p-values.

Explore simple linear regression with an intuitive example, derive the ordinary least squares equation, and use Excel to estimate slope, correlation coefficient, and coefficient of determination.

Explore how self-study hours relate to math scores using a scatter plot, showing a general linear trend and how the least squares method derives the linear relationship.

Apply the ordinary least squares method to estimate y = a + b x by minimizing squared errors; a = ȳ − b x̄ and b = Σ(X−x̄)(Y−ȳ)/Σ(X−x̄)².

Explore how Excel's add trendline feature computes the regression line via ordinary least squares, displaying the slope, intercept, and chart-based equations for a scatter plot.

Learn how the ordinary least squares method creates the regression line and uses R-squared, the coefficient of determination, to measure the strength of linear relationships.

Learn Pearson's r, the correlation coefficient that quantifies the strength and direction of a linear relationship, with values from -1 to 1 and stronger relationships near the extremes.

Learn how linear regression reveals the relationship between X and Y to predict Y from X, including multiple linear regression with one dependent variable and multiple independent variables.

Learn to perform linear regression using the Excel data analysis procedure, then manually calculate and validate the regression coefficient, intercept, slope, R, and R-square with residuals.

Explore categorical data analysis, describe two-way relationships between variables using the chi-square statistic, its formula and degrees of freedom, and apply Excel's chi-square test of independence.

Identify and classify data as categorical, contrasting text-based categories such as state, gender, and country, with coded numeric categories like income group and age group, to learn proper data treatment.

Describe categorical data using frequency distributions, pivot tables, and visualizations like bar and pie charts. Learn to assess two-way interactions with cross-tab techniques.

Describe contingency tables and cross-tabs for two-way categorical data, showing counts and percent interactions between two variables, and introduce chi-square to compare association strength.

Explore the chi-square statistic for independence, comparing observed and expected counts in contingency tables to assess how gender, region, or education influence exam success rates.

Explore the chi-square statistic by analyzing observed minus expected squared over expected, noting how squaring magnifies deviations and makes them positive, with applications to contingency tables and degrees of freedom.

Explore degree of freedom in contingency tables and its role in chi-square significance. Learn the formula df = (R-1)(C-1) and how to compare chi-square to critical values.

Explain the chi-square distribution and the statistic (observed minus expected) squared over expected, then show how degrees of freedom set the mean, variance, skew, and critical values.

Learn how to perform a chi-square test in Excel, compare p-values across degrees of freedom, compute observed and expected counts, and identify which variable more strongly predicts the outcome.

Distinguish dependent (outcome) and independent (explanatory) variables with regression examples, like hours of study versus score and height versus weight, and learn how to interpret their relationships using chi-square.

A practical guide shows when to use chi-square, linear regression, ANOVA, or logistic regression based on the dependent and independent variable types (categorical or numeric).

Learn when to use ANOVA, including one-way ANOVA, its hypotheses and assumptions, and how to compute within and between variance, F-statistics and F-distribution, with Excel demonstrations.

Explore scenarios for one-way ANOVA across distinct groups, comparing means of scores, mileage, and bacteria counts to determine if group differences are due to treatment rather than chance.

Explore the design of ANOVA for comparing more than two group means with a categorical independent variable and dependent variable, including within and between variance and null and alternative hypotheses.

Explore the assumptions of ANOVA, including zero-mean errors, equal variances, and independence, and learn to check normality with histograms and Q-Q plots, plus a 2:1 variance rule.

Compare three samples using between-sample variance and within-sample variance to assess if they come from the same population, using the F-statistic and related degrees of freedom.

Compute between sample variance by weighting each group's mean deviation from the grand mean with its sample size, then divide by the number of groups minus one, as shown in Excel.

Learn to calculate within-sample variance across multiple groups, weighting each variance by n_i−1, and obtain the pooled estimate of population variance with degrees of freedom N−k.

The demonstration shows that when samples come from the same population, the between and within variance yield an F-statistic near one, validating the null hypothesis.

Analyze how samples from different populations affect variance: within-sample variance stays similar while between-sample variance grows when one sample is very different, yielding large f-statistics and supporting the alternate hypothesis.

Explore the F-statistics and F-distribution, and see from an example how between and within variance form the F ratio with degrees of freedom and how the F-table signals significance.

Learn to run one-way ANOVA in Excel using the data analysis toolpak. The method computes between and within variances, F-statistics, and p-values, with clear interpretation at the 5 percent level.

Explore non-parametric tests in business contexts, compare monotony and linearity, and apply rank-based tests like rank-sum, Mann-Whitney, Kruskal-Wallis, run tests, and Spearman correlation with practical examples.

Explore when to use non-parametric tests with ordinal data, ranks, and distribution-free assumptions, including real-life surveys and rank comparisons in physics and maths.

Understand the difference between monotony and linearity, and how monotonic versus linear relationships affect regression and non-parametric methods, with examples like y = a + b x.

Non-parametric methods offer distribution-free analysis for ordinal and nominal data with simple calculations and robustness, but they trade off lower power and larger sample needs.

Apply the non-parametric sign test to paired data by counting plus and minus changes, ignoring ties. Compare the observed proportion to 0.5 using normal approximation to assess training impact.

Learn the Mann Whitney U test, a non-parametric rank-sum method to compare two independent samples and test if they come from the same population, using rank sums and U statistic.

Learn the Kruskal-Wallis H test for three independent samples, without distribution assumptions, using rank sums and a chi-square comparison, shown with video tutorials, study groups, and classroom training.

Explore the one sample run test, a non-parametric method to assess randomness in a two-type pattern, such as male versus female, using runs, mean, and standard deviation.

Explain Spearman's rank correlation for ordinal data, measuring rank association strength and direction. Demonstrate with aptitude versus final exam ranks to compute rho and identify strong or perfect correlations.

Finish by highlighting mastery of statistics basics, p-values, and type i and type ii errors, while inviting feedback and pointing to related courses like logistic regression and cluster analysis.