Bivariate distributions Revision Notes

In statistical studies, variables represent characteristics or quantities that can change or vary. When examining relationships between two variables, it is important to distinguish their roles to understand how one may influence the other. This distinction helps in setting up experiments and analysing results effectively.

The independent variable is the factor that is deliberately changed or controlled by the researcher to observe its effect. It is sometimes called the explanatory or predictor variable.

The dependent variable is the outcome or response that is measured to see how it changes when the independent variable is varied. It depends on the independent variable.

For example, if you want to study how the amount of sunlight affects plant growth, the amount of sunlight is the independent variable because you control it, and the plant growth (such as height) is the dependent variable because it responds to the sunlight levels.

• Independent variable: changed or controlled by the experimenter.
• Dependent variable: measured or observed outcome influenced by the independent variable.

To identify which variable is independent or dependent, consider which factor you can control or select and which factor you measure as a result.

Example 1: Investigating how study time affects test scores. Here, study time is independent (you choose how long to study), and test score is dependent (it depends on study time).

Example 2: Examining the relationship between temperature and ice cream sales. Temperature is independent (weather conditions), and ice cream sales are dependent (they vary with temperature).

• Ask: Which variable do I change or select? That is independent.
• Ask: Which variable do I measure or observe? That is dependent.

Correctly identifying independent and dependent variables is crucial for designing valid experiments and interpreting results.

It ensures that the cause-effect relationship is clear and that data collection focuses on measuring the correct outcomes.

Misidentifying variables can lead to incorrect conclusions about relationships between variables.

• Helps establish cause and effect in experiments.
• Guides data collection and analysis.
• Prevents confusion in interpreting results.

Remember that the independent variable is what you change or control, and the dependent variable is what you measure.

In observational studies where variables are not controlled, the independent variable is often the one considered as the predictor or explanatory variable.

Always state variables clearly when describing a study to avoid ambiguity.

• Independent variable: controlled or selected factor.
• Dependent variable: measured response.
• Clear identification aids understanding and communication of results.

Correlation is a statistical concept that describes the relationship between two variables and how they change together. It tells us whether an increase in one variable tends to be associated with an increase or decrease in the other variable, and how consistent this relationship is.

For example, consider the relationship between hours studied and exam scores. If students who study more hours tend to get higher scores, these two variables show a positive correlation. Correlation does not imply causation but simply describes the pattern of association between variables.

• Correlation measures the direction and strength of a relationship between two variables.
• It can be observed visually using scatter diagrams or numerically using tables of paired data.

Correlation can be positive or negative depending on the direction of the relationship between variables.

Positive correlation means that as one variable increases, the other variable also tends to increase.

Negative correlation means that as one variable increases, the other variable tends to decrease.

• Positive correlation example: Height and weight – generally, taller people weigh more.
• Negative correlation example: Number of hours spent watching TV and exam scores – more TV time might be associated with lower scores.

The strength of correlation describes how closely the data points follow a clear pattern.

Strong correlation means the points lie close to a straight line, showing a consistent relationship.

Weak correlation means the points are more spread out, indicating a less consistent relationship between variables.

• Strong correlation example: Temperature and ice cream sales – points cluster tightly along an upward trend.
• Weak correlation example: Age and number of text messages sent – points show a general trend but with much scatter.

No correlation means there is no apparent relationship between the two variables. The points on a scatter diagram appear randomly scattered without any clear pattern or trend.

For example, shoe size and exam score typically show no correlation because knowing one does not help predict the other.

• No correlation example: Shoe size and exam score – points scattered without any trend.

Scatter diagrams plot paired data points on a graph with one variable on the x-axis and the other on the y-axis. By observing the pattern of points, you can identify the type of correlation.

Tables of paired data can also be examined by looking at how values of one variable change relative to the other.

Steps to identify correlation from a scatter diagram: 1. Look for an overall trend (upwards, downwards, or none). 2. Assess how closely points cluster around a line (tight or scattered). 3. Decide if the correlation is positive, negative, strong, weak, or none.

• Scatter diagrams visually show the relationship between two variables.
• Tables allow comparison of paired values to detect trends.
• Identifying correlation involves assessing direction and strength from data patterns.

Bivariate data consist of pairs of related numerical values collected from the same subjects or units. For example, the heights and weights of students form bivariate data because each height corresponds to a specific weight.

A scatter diagram is a graph that displays these pairs as points on a coordinate plane, allowing visual examination of any relationship between the two variables. Unlike single-variable graphs, scatter diagrams show how one variable may change in relation to the other.

• Bivariate data = paired numerical values from the same subjects.
• Scatter diagrams plot these pairs as points on two axes.
• They help identify relationships or patterns between variables.

To create a scatter diagram, first decide which variable is independent (often the cause or input) and which is dependent (often the effect or output). The independent variable is plotted on the horizontal axis (x-axis), and the dependent variable on the vertical axis (y-axis).

Label each axis clearly with the variable name and units if applicable. Choose an appropriate scale that covers all data values evenly to avoid distortion.

For each pair of values, locate the x-coordinate on the horizontal axis and the y-coordinate on the vertical axis, then mark a point where these meet. Repeat for all data pairs to complete the diagram.

• Identify independent (x) and dependent (y) variables.
• Label axes with variable names and units.
• Choose scales that include all data points.
• Plot each data pair as a point (x, y).

Suppose we record the number of hours five students studied and their corresponding test scores: (2 hours, 65 marks), (4 hours, 75 marks), (3 hours, 70 marks), (5 hours, 80 marks), (1 hour, 60 marks).

Here, study hours is the independent variable (x-axis) and test score is the dependent variable (y-axis). After labelling axes and choosing scales (e.g., 0 to 6 hours on x-axis, 50 to 90 marks on y-axis), plot each point accordingly.

This scatter diagram visually shows how test scores tend to increase as study hours increase, suggesting a positive relationship.

• Data pairs: (2,65), (4,75), (3,70), (5,80), (1,60).
• Plot hours studied on x-axis, test scores on y-axis.
• Each point represents one student's data.
• Pattern suggests positive correlation.

Once all points are plotted, observe the overall pattern. Points clustered along an upward slope indicate positive correlation, where increases in one variable relate to increases in the other.

A downward slope suggests negative correlation, where one variable increases as the other decreases. If points are scattered without any clear pattern, this indicates no correlation.

Scatter diagrams do not prove causation but help identify associations worth further investigation.

• Upward trend = positive correlation.
• Downward trend = negative correlation.
• No clear pattern = no correlation.
• Scatter diagrams suggest associations, not causation.

Ensure accuracy when plotting points by using graph paper or digital tools with precise scales. Avoid overcrowding points by choosing suitable scales and graph sizes.

Scatter diagrams are limited to two variables; more complex relationships require other methods.

Remember that outliers (points far from the main cluster) can affect interpretation and should be noted for further analysis.

• Use precise scales and tools for accuracy.
• Scatter diagrams show only two variables at once.
• Identify and consider outliers carefully.

A scatter diagram is a graph that displays pairs of data points, each representing two related variables. For example, plotting hours studied (x) against exam scores (y) for several students shows how these variables might be related.

A line of best fit is a straight line drawn on the scatter diagram that summarises the overall trend of the data points. It helps to identify whether there is a positive, negative, or no correlation between the variables and can be used to estimate values within the data range.

• Scatter diagrams plot paired data points (x, y).
• The line of best fit represents the general trend of the data.
• It balances the data points to minimise overall distance from the line.

Visual estimation involves looking at the scatter diagram and drawing a straight line that best represents the trend of the points. The goal is to position the line so that roughly half the points lie above it and half below, minimising the total vertical distances from the points to the line.

This method is quick and useful when an exact calculation is not required, but it depends on careful observation and practice to avoid bias.

For example, if a scatter diagram shows points rising from left to right, the line of best fit should slope upwards, passing through the middle of the cluster of points.

• Draw a straight line through the data that balances points above and below.
• Aim to minimise the sum of vertical distances from points to the line.
• Useful for quick, approximate trend identification.

The method of semi-averages provides a more systematic way to draw the line of best fit. It requires an even number of data pairs.

First, order the data pairs by the independent variable (usually x) from smallest to largest.

Then split the data into two equal halves. For each half, calculate the average x-value and the average y-value. These averages give two points: (x̄₁, ȳ₁) for the first half and (x̄₂, ȳ₂) for the second half.

• Ensure the number of data pairs is even.
• Order data by x-values and split into two equal halves.
• Calculate average x and y for each half.
• Plot the two average points on the scatter diagram.

After plotting the two average points, draw a straight line through them. This line is the line of best fit according to the semi-averages method.

For example, consider the data pairs: (1,2), (2,3), (3,5), (4,4).

Split into two halves: first half (1,2), (2,3); second half (3,5), (4,4). Calculate averages: first half x̄₁ = (1+2)/2 = 1.5, ȳ₁ = (2+3)/2 = 2.5; second half x̄₂ = (3+4)/2 = 3.5, ȳ₂ = (5+4)/2 = 4.5. Plot points (1.5, 2.5) and (3.5, 4.5) and draw the line through them. This line summarises the trend in the data and can be used to estimate y for values of x between 1 and 4.

• Draw a line through the two average points.
• Use the line to interpret the relationship between variables.
• Apply the line for predictions within the data range (interpolation).

The line of best fit summarises the relationship between the two variables. If the line slopes upwards, it indicates a positive correlation; if it slopes downwards, a negative correlation.

You can use the line to estimate the value of one variable given the other by reading values from the graph or calculating from the line equation if known.

Be cautious when making predictions outside the range of the data (extrapolation), as the relationship may not hold beyond observed values.

• Slope direction indicates correlation type.
• Use the line to estimate unknown values (interpolation).
• Avoid extrapolation beyond the data range without justification.

When analysing two related variables, a scatter diagram shows how one variable changes with the other. The line of best fit is a straight line drawn through the scatter plot that best represents the overall trend of the data points. It summarises the relationship between the variables and helps predict values of one variable based on the other.

• Summarises the trend in bivariate data.
• Helps identify positive, negative, or no correlation.
• Used for estimation and interpretation.

The gradient of the line of best fit indicates the rate at which the dependent variable (y) changes for each unit increase in the independent variable (x). To calculate the gradient, select two points on the line (not necessarily data points) and use the formula:

\n\nm = (change in y) / (change in x) = (y₂ - y₁) / (x₂ - x₁)\n\n

For example, if the line passes through points (2, 5) and (6, 13), then: m = (13 - 5) / (6 - 2) = 8 / 4 = 2. This means y increases by 2 units for every 1 unit increase in x.

• Gradient shows rate of change of y with respect to x.
• Use two points on the line, not necessarily original data points.
• Calculate as rise over run.

The y-intercept is the value of y when x = 0. It is found by identifying where the line of best fit crosses the y-axis. If the line does not cross exactly at a plotted point, estimate the intercept by reading the graph carefully or use the equation y = mx + c with a known point to solve for c.

For example, using the point (2, 5) and gradient m = 2:

\n\n5 = 2 × 2 + c → 5 = 4 + c → c = 1\n\nSo, the y-intercept is 1.

• Y-intercept is the value of y when x = 0.
• Can be read directly from the graph or calculated using a point and the gradient.
• Represents the starting value of y before x changes.

Once the gradient and y-intercept are known, write the equation of the line of best fit in the form y = mx + c. This equation models the relationship between x and y.

Using the previous example, with m = 2 and c = 1, the equation is: y = 2x + 1.

This equation can be used to estimate y for any given x within the range of the data (interpolation).

• Equation summarises the linear relationship.
• Use calculated m and c values.
• Essential for prediction and analysis.

The line of best fit equation allows estimation of y values for given x values that may not be in the original data set. For example, if x = 4, then y = 2 × 4 + 1 = 9.

Note that predictions are most reliable within the range of observed data (interpolation). Extrapolation beyond this range can be inaccurate because the linear trend may not continue.

Always state units and context when interpreting predictions.

• Estimate y for given x using the equation.
• Prefer interpolation over extrapolation for accuracy.
• Interpret results in context with units.

A line of best fit is a straight line drawn through a scatter diagram that best represents the trend of the data points. It summarises the relationship between two variables, often called the independent variable (x) and the dependent variable (y). The line minimises the overall distance of points from itself, showing the general direction of the association, whether positive, negative, or none.

• It captures the average pattern in the data.
• Helps to identify the nature of the relationship between variables.
• Forms the basis for estimating unknown values.

To estimate the value of one variable from the other using the graph: first locate the known value on the horizontal axis (usually x). From this point, move vertically until you meet the line of best fit. Then, move horizontally to the vertical axis (usually y) and read off the estimated value. This process is called interpolation when the known value lies within the range of the data.

For example, if a scatter plot shows hours studied (x) against test score (y), and you want to estimate the test score for 5 hours of study, find 5 on the x-axis, move up to the line, then across to the y-axis to read the predicted score.

• Use the horizontal axis to find the known variable.
• Move vertically to the line of best fit.
• Move horizontally to read the estimated value on the other axis.
• Best used for values within the data range (interpolation).

If the equation of the line of best fit is known, usually in the form y = mx + c, you can estimate the unknown variable by substituting the known value into the equation. Here, m is the gradient (rate of change) and c is the y-intercept (value of y when x = 0).

For example, if the line is y = 3x + 2 and you want to estimate y when x = 4, substitute x = 4: y = 3(4) + 2 = 12 + 2 = 14. So, the estimated value of y is 14.

• Identify the equation of the line of best fit.
• Substitute the known value into the equation.
• Calculate the result to find the estimate.
• Check units and context for interpretation.

Using a line of best fit assumes a linear relationship between the variables. Estimates are most reliable when the known value lies within the range of the original data (interpolation). Estimating values outside this range (extrapolation) can be unreliable because the linear trend may not continue.

Also, the line of best fit summarises average behaviour; individual data points may vary, so estimates are approximate. Always consider the scatter of points around the line and the strength of correlation before trusting estimates.

• Assumes linear relationship between variables.
• Reliable mainly for interpolation within data range.
• Extrapolation beyond data range is uncertain.
• Estimates are approximate, not exact.

A scatter diagram shows the relationship between hours of revision (x) and exam score (y). The line of best fit has equation y = 5x + 40. Estimate the exam score for a student who revises for 6 hours.

Substitute x = 6 into the equation: y = 5(6) + 40 = 30 + 40 = 70.

Therefore, the estimated exam score is 70. This estimate assumes the linear trend holds and 6 hours is within the data range.

• Equation: y = 5x + 40
• Known x: 6 hours
• Calculation: y = 5 × 6 + 40 = 70
• Interpretation: Estimated exam score is 70.

A line of best fit is a straight line drawn through a scatter diagram to model the relationship between two variables, usually denoted as x (independent variable) and y (dependent variable). It summarises the trend in the data and can be used to estimate y for given values of x.

The line is typically found by methods such as semi-averages or least squares, aiming to minimise the overall distance between the data points and the line itself.

• Models a linear relationship between two variables.
• Used to estimate values of one variable based on the other.

Several assumptions underpin the use of a line of best fit for estimation:

1. The relationship between the variables is linear, meaning the data points roughly follow a straight-line pattern.

2. The pattern of the relationship observed in the data continues consistently within the range of the data points (interpolation). This means the line accurately represents the association between variables in this range, allowing reliable estimation of unknown values within it.

• Linear relationship between variables.
• Consistent pattern within the observed data range.
• Data points are independent of each other.
• Random variation around the line is minimal.

Interpolation means estimating values of y for x-values that lie within the range of the observed data. This is generally reliable because the line of best fit is based on actual data points in this range.

Extrapolation means estimating values of y for x-values outside the observed data range. This is risky because the linear trend may not continue beyond the data collected. The relationship could change, become non-linear, or break down entirely, leading to inaccurate or misleading estimates.

For example, if data is collected for temperatures between 10°C and 30°C, estimating y at 25°C (interpolation) is more reliable than estimating at 40°C (extrapolation).

• Interpolation: estimation within data range, generally reliable.
• Extrapolation: estimation outside data range, potentially unreliable.

Suppose a line of best fit for data relating hours studied (x) to exam score (y) is given by y = 5x + 40.

To estimate the exam score for a student who studied 6 hours (within the data range 0 to 8 hours):

Substitute x = 6 into the equation: y = 5(6) + 40 = 30 + 40 = 70. So, the estimated score is 70%. This is interpolation and likely reliable if 6 hours lies within the observed data range.

• Use the equation y = mx + c for estimation.
• Substitute the known x-value to find estimated y.
• Check that x-value lies within the observed data range for reliability.

The accuracy of estimates depends on the assumptions holding true. If the data points are not independent (for example, repeated measurements from the same subject without accounting for this), the line may not represent the true relationship.

Random variation or outliers can affect the line of best fit, reducing estimation accuracy.

Extrapolation can lead to large errors because the relationship may change outside the observed range. Always be cautious and justify any extrapolation used in answers.

• Data independence is important for valid estimation.
• Outliers can distort the line of best fit.
• Avoid or clearly justify extrapolation beyond data range.