Back to Quantitative Analysis
KASNEB · FoundationQuantitative AnalysisBETA — flag if wrong

Regression Analysis

This topic covers the principles and applications of regression analysis in predicting outcomes based on data.

3objectives
3revision lessons
12practice questions

What you’ll learn

Aligned to the KASNEB Quantitative Analysis syllabus.

Understanding Regression Analysis in Quantitative Research

BETA — flag if wrongAI 100

Regression analysis is a statistical method used to examine the relationship between one dependent variable and one or more independent variables. Its primary purpose is to model and analyze the relationships to predict outcomes, identify trends, and inform decision-making processes in various fields, including finance, marketing, and economics.

In quantitative research, regression analysis helps in understanding how changes in independent variables affect the dependent variable. For instance, a business may want to predict sales based on advertising expenditure and market conditions. By applying regression analysis, the business can quantify the impact of each factor and make informed decisions on resource allocation.

There are different types of regression, including linear regression, which assumes a straight-line relationship, and multiple regression, which involves multiple independent variables. The output of regression analysis includes coefficients that indicate the strength and direction of the relationships, along with statistical significance tests to validate the findings.

In the Kenyan context, businesses can leverage regression analysis to optimize pricing strategies, forecast demand, and assess the impact of economic policies. Tools like Excel or statistical software can facilitate the analysis, making it accessible for various organizations, from SMEs to large corporations.

Key points

  • Regression analysis models relationships between variables.
  • It predicts outcomes and identifies trends in data.
  • Linear and multiple regression are common types.
  • Coefficients indicate strength and direction of relationships.
  • Useful for decision-making in Kenyan businesses.
Worked example

Consider a company that wants to predict its sales (KES) based on advertising spend (KES). The data collected is as follows:

| Advertising Spend (KES) | Sales (KES) | |-------------------------|-------------| | 10,000 | 100,000 | | 20,000 | 150,000 | | 30,000 | 200,000 | | 40,000 | 250,000 |

Using linear regression, we can derive the equation of the line: Sales = a + b * Advertising Spend.

  1. Calculate the coefficients (a and b) using the least squares method. Suppose we find:

    • b (slope) = 5
    • a (intercept) = 50,000

  2. The regression equation becomes: Sales = 50,000 + 5 * Advertising Spend.

  3. To predict sales for an advertising spend of KES 25,000: Sales = 50,000 + 5 * 25,000 = 50,000 + 125,000 = 175,000 KES.

Thus, the predicted sales for KES 25,000 in advertising is KES 175,000.

More on this topic

CF12.8.B Computing Simple Linear Regression CoefficientsBETA — flag if wrongAI 100
Simple linear regression analyzes the relationship between two variables by fitting a linear equation to observed data. The formula for the regression line is:

\[ Y = a + bX \]

Where:
- Y = dependent variable (the outcome)
- X = independent variable (the predictor)
- a = Y-intercept (the value of Y when X = 0)
- b = slope of the line (the change in Y for a one-unit change in X)

To compute the coefficients:
1. Calculate the means of X and Y.
2. Determine the slope (b) using the formula:
\[ b = \frac{\sum{(X_i - \bar{X})(Y_i - \bar{Y})}}{\sum{(X_i - \bar{X})^2}} \]
3. Calculate the Y-intercept (a) using:
\[ a = \bar{Y} - b\bar{X} \]
4. Interpret the coefficients: The slope indicates how much Y changes for each unit increase in X. The intercept shows the expected value of Y when X is zero.

In Kenya, businesses use regression analysis for forecasting sales, understanding market trends, and making data-driven decisions.
CF12.8.C Evaluating Regression Models for ForecastingBETA — flag if wrongAI 100
Regression analysis is a statistical method used to model and analyze the relationships between variables. It helps in forecasting future values based on historical data. In Kenya, businesses utilize regression analysis for various purposes, such as predicting sales, expenses, or market trends. The effectiveness of regression models can be evaluated using several key metrics:

1. R-squared (R²): This statistic indicates the proportion of variance in the dependent variable that can be explained by the independent variables. An R² value closer to 1 suggests a strong model fit.

2. Adjusted R-squared: Unlike R², this metric adjusts for the number of predictors in the model, providing a more accurate measure when comparing models with different numbers of independent variables.

3. Standard Error of Estimate: This measures the average distance that the observed values fall from the regression line. A smaller standard error indicates a better fit.

4. P-values: These help determine the significance of individual predictors. A p-value less than 0.05 typically indicates that the predictor is statistically significant.

5. Residual Analysis: Examining the residuals (the differences between observed and predicted values) helps identify patterns that may suggest model inadequacies or violations of regression assumptions.

In practice, businesses in Kenya can apply these metrics to refine their forecasting models, ensuring they remain relevant and accurate in a dynamic market environment.

Sample KASNEB-style questions

3 of 12 questions. Beta-flagged questions are AI-drafted and pending CPA review — flag anything that looks wrong.

Q1 · SHORT ANSWER · hardBETA — flag if wrong

An insurance company has 2 claim assessors who must each give approval to customers who wish to lodge a claim for compensation. The manager currently has 8 such customers and has asked each assessor to independently rank the customer claims in order of merit. The rankings are shown below: Customer A B C D E F G H Assessor 1 Ranking 4 6 2 1 5 8 6 3 Assessor 2 Ranking 4 8 1 3 7 4 6 2 Required: Calculate the rank correlation coefficient of the two sets of rankings and comment on the results. (8 marks)

Model answer

The rank correlation coefficient of the two sets Since there are tired rankings: r = 1 - [ ( ) ] d = or Where: = Assessor 1 ranking = Assessor 2 ranking d = 4 4 4.5 -0.5 0.25 6 6.5 8 -1.5 2.25 2 1 1 1 1 3 -2 4 5 7 -2 4 8 4 4.5 3.5 12.25 6 6 0.5 0.25 3 2 1 1 = 25 t = 2 for and t = 2 for r = 1 - [ ] = 1- = 0.6905 Approx. Hence, there is a strong positive correlation between the ranking of the customer claims in order of merit made by the two assessors.

Q2 · SHORT ANSWER · easyBETA — flag if wrong

An accountant wishes to undertake a cost analysis of the annual repair cost for a popular model of a machine as influenced by the age of the machine. The results obtained are shown below: Age (years) Repair cost (Sh.“000”) 1 70 3 140 5 230 8 350 7 300 12 570 8 350 4 200 Required: (i) Pearson‘s coefficient of correlation between the age of the machine and the repair cost. Interpret your result. (4 marks) (ii) Fit a least squares regression line of repair cost on age of machine to the data. (4 marks) (iii) Interpret the meaning of regression coefficients a and b in the least squares regression line obtained in (c) (ii) above. (2 marks) (iv) The coefficient of determination. Interpret your result. (2 marks)

Model answer

(i) Pearson‟s coefficient of correlation r = √[ ] [ ]] Where: x = Age (years) y = Repair cost (sh. ― 000‖) x y xy 𝒚 1 70 70 1 4900 3 140 420 9 19600 5 230 1150 25 52900 8 350 2800 64 122500 7 300 2100 49 90000 12 570 6840 144 324900 8 350 2800 64 122500 4 200 800 16 40000 48 2210 16980 372 777300 n = 8 r = [ ] [ ] = √ = 0.9939. Hence, there‘s a strong positive correlation between the age of the machine and the repair cost. (ii) Least squares regression line Y = a+bx a = b = b = = sh. 44.286 per year. a = – 44.286 = 10.534 (sh.000) y = 10.534 +44.286x (iii) Interpretation ― a‖ represents the repair cost of the machine that is independent of the age of the machine. On the other hand, ― b‖ represents that part of the repair cost of the machine that it dependent on the age of the machine i.e as the machine ages it costs more to repair as vice versa. (iv) The coefficient of determination = (r r = 0.9939 = (0.9939 = 0.9878 or 98.78%. Hence, about 98.78% of the variability in the repair cost of the machine is accounted for by the age of the machine.

Q3 · SHORT ANSWER · easyBETA — flag if wrong

In a choral music competition, 9 contestants were awarded marks in percentage using a music scoring grid by two assessors. The results obtained were given as shown in the table below: Marks in % by: Contestant 1st Assessor 2nd Assessor A 72 76 B 82 80 C 79 78 D 70 73 E 67 70 F 81 85 G 78 69 H 75 83 I 65 68 Required: (i) The rank correlation coefficient. Interpret your results. (4 marks) (ii) Coefficient of determination. (1 mark)

Model answer

Correlation and regression analysis i) Rank correlation coefficient Contestants 1st Assessor Rank 2nd Assessor Rank d A 72 6 76 5 1 1 B 82 1 80 3 -2 4 C 79 3 78 4 -1 1 D 70 7 73 6 1 1 E 67 8 70 7 1 1 F 81 2 85 1 1 1 G 78 4 69 8 -4 16 H 75 5 83 2 3 9 I 65 9 68 9 0 0 34 = 1 – 0.283 = 0.7167 ii) Coefficient of determination = = 0.5137

Practice the full question bank with the AI tutor

12 questions on this topic alone. Get feedback after every attempt; the tutor re-explains what you got wrong. Beta access is free.

Reserve beta access

Common questions

Define regression analysis and its purpose in quantitative research.

Regression analysis models relationships between variables.

Compute and interpret simple linear regression coefficients.

Simple linear regression fits a line to data points.

Evaluate the effectiveness of regression models in forecasting.

R-squared indicates the variance explained by the model.

More from Quantitative Analysis

Introduction to Quantitative Analysis

Topic 1

Data Collection Methods

Topic 2

Descriptive Statistics

Topic 3

Probability Theory

Topic 4

Distribution Theory

Topic 5

Sampling Techniques

Topic 6

Hypothesis TestingForecasting Techniques

AI tutor for the full CPA pathway

Quantitative Analysis is one of 18 CPA papers covered. Beta access is free; KES 1,500/month at launch.

See the full CPA pathway →