CAIIB Paper 1 Module A Quiz 7

Systematic sampling selects every kth unit after a random start. It is efficient, although periodic patterns in the list can introduce bias.

CV = standard deviation / mean × 100. A: 3/12 = 25%; B: 6/15 = 40%. Portfolio A has lower relative variability by this measure.

Trend captures the underlying long-term direction of a time series, distinct from seasonal, cyclical and irregular fluctuations.

The Poisson distribution is commonly used for counts of events in a fixed interval when events occur independently at a stable average rate under suitable assumptions.

Constraints translate resource limitations or required conditions into linear equations or inequalities.

Data cleaning and validation should precede analysis. Duplicates can inflate counts or sums, while missing values may distort estimates if handled carelessly.

The analyst should check definitions, coverage, collection methods, dates and reliability across sources. Combining incompatible data can create misleading comparisons.

A negative slope means the predicted dependent variable decreases by 3 units for each one-unit increase in the predictor, on average in the fitted linear model.

Positive slack indicates the constraint is not fully used at that solution; some of the resource remains unused.

Expected return describes the average outcome, while standard deviation describes dispersion around the mean and therefore one aspect of risk.

In cluster sampling, groups such as branches are sampled, and all units or a sample of units within selected clusters are observed.

A long right tail indicates positive (right) skewness. In a typical positively skewed distribution, mean exceeds median.

Cyclical variation relates to business-cycle movements and typically spans more than one year, unlike regular within-year seasonality.

The normal distribution is symmetric and unimodal, so its mean, median and mode coincide at the centre.

Decision variables represent the choices under the decision-maker‘s control; the objective and constraints are expressed in terms of them.

Using class boundaries 9.5–19.5, the width is 10. Equivalently, adjacent class limits differ by 10.

An open-ended class lacks a known upper boundary and midpoint, so estimating a grouped mean requires an additional assumption or method.

Ordinary least squares minimises the sum of squared residuals, not the sum of absolute residuals or the sum of predictions.

Slack variables convert ≤ constraints into equalities and can provide an initial basic feasible solution in a standard-form simplex setup when the conditions allow.

The conditional proportion is 20/100 = 0.20, or 20%. This is the observed P(fraud|flag) for the sample of flagged transactions.