Physiology Lab Data
Graphs, Trends, and Error Bars
Learn to read, interpret, and critically analyze physiological data including graph types, trend lines, and the role of error bars in experimental results.
Anatomy & Physiology
What Is Lab Data?
In physiology experiments, we collect quantitative measurements: heart rates, blood pressures, reaction times, and enzyme activities. Raw numbers alone are hard to interpret, so we transform them into graphs and summary statistics that reveal patterns.
Independent Variable
The factor you deliberately change (e.g., exercise intensity)
Dependent Variable
The outcome you measure (e.g., heart rate in bpm)
Controlled Variables
Factors kept constant to ensure a fair test
Common Graph Types in Physiology
Line Graph
Shows continuous change over time or across conditions: heart rate during exercise
Bar Graph
Compares discrete categories: average blood pressure at rest vs. after caffeine
Scatter Plot
Shows correlation between two continuous variables: BMI vs. resting heart rate
Anatomy of a Physiology Graph
What Are Error Bars?
Error bars are graphical representations of variability in data. They extend above and below each data point to show the range within which the true population value likely falls.
Standard Deviation (SD)
Shows spread of individual measurements around the mean
Standard Error (SE)
Shows precision of the sample mean as an estimate of the population mean
95% Confidence Interval
Range within which we are 95% confident the true mean lies
SD vs. SE vs. 95% CI
| Standard Deviation | Standard Error | 95% CI | |
|---|---|---|---|
| What it shows | Spread of individual data | Precision of the mean | Range for the true mean |
| Formula | \( s \) | \( \frac{s}{\sqrt{n}} \) | \( \bar{x} \pm 1.96 \cdot \text{SE} \) |
| Bar size | Largest | Smaller | Moderate |
| Affected by \( n \)? | Not directly | Yes: shrinks as \( n \) grows | Yes: narrows as \( n \) grows |
| Best for | Describing variability | Comparing group means | Statistical inference |
Interpreting Trends in Data
Positive Trend
As stimulus increases, response increases: e.g., exercise intensity and heart rate
Plateau
Response levels off despite increasing stimulus: e.g., enzyme saturation
Inverse Trend
As stimulus increases, response decreases: e.g., parasympathetic tone and heart rate
Overlapping Error Bars and Significance
Reading Bar Graphs with Error Bars
1
Identify the axes: What is the independent variable (x-axis) and the dependent variable (y-axis)?
2
Read the bar heights: Each bar represents the mean for that group or condition.
3
Examine error bars: Note which type (SD, SE, or CI) and whether they overlap between groups.
4
Judge significance: Non-overlapping SE or CI bars strongly suggest a statistically significant difference.
5
Look for patterns: Compare across all groups, not just two at a time.
Example 1: Caffeine and Heart Rate
Example 2: Scatter Plot and Correlation
Example 3: Do These Groups Differ?
How to Analyze Lab Data
1
Read the title and axes. Identify variables, units, and the type of graph.
2
Note the scale. Check if the y-axis starts at zero. A truncated axis can exaggerate small differences.
3
Read the means. Compare bar heights or data point positions across groups.
4
Check error bar type. Look for a legend or caption: is it SD, SE, or 95% CI?
5
Compare error bars. Overlap suggests no significant difference; a clear gap supports significance.
6
Describe the trend. Positive, negative, plateau, or no clear pattern? State it in a conclusion.
Sample Size and Error Bar Width
Common Mistakes
✕ Confusing SD with SE
SD describes spread in the data; SE describes precision of the mean. Using SD when you mean SE inflates apparent variability.
✕ Ignoring error bars
Two bars may look different in height, but if their error bars overlap heavily, the difference may not be real.
✕ Assuming causation
A scatter plot correlation does not mean one variable causes the other. Always consider confounding variables.
Summary of Key Concepts
| Concept | Key Point | |
|---|---|---|
| Error bars (SE) | Precision of the mean | Smaller with larger \( n \) |
| Error bars (SD) | Spread of individual data | Stable regardless of \( n \) |
| 95% CI | Range for the true mean | \( \bar{x} \pm 1.96 \cdot \text{SE} \) |
| Overlap rule | Overlap = likely not significant | No overlap = likely significant |
| Correlation (\( r \)) | Strength of linear relationship | \( |r| \) near 1 = strong |
| Causation | Requires controlled experiment | Correlation alone is insufficient |
Question 1 of 5
A bar graph shows mean blood glucose levels for three groups. Group A has an error bar (SE) from 90 to 110 mg/dL, and Group B has an error bar from 105 to 125 mg/dL. What can you conclude?
The SE bars overlap (Group A's upper limit of 110 falls within Group B's range of 105-125). Overlapping SE bars suggest the difference between means is likely not statistically significant.
Question 2 of 5
Which formula correctly calculates the standard error of the mean?
The standard error is calculated as \( \text{SE} = \frac{s}{\sqrt{n}} \), where \( s \) is the standard deviation and \( n \) is the sample size. Dividing by \( \sqrt{n} \) means larger samples yield smaller SE values.
Question 3 of 5
A scatter plot of exercise frequency vs. resting heart rate shows \( r = -0.82 \). Which interpretation is most accurate?
An \( r \) value of \( -0.82 \) indicates a strong negative correlation. However, correlation alone does not prove causation: a controlled experiment would be needed to establish that exercise causes lower HR.
Question 4 of 5
A researcher increases the sample size from \( n = 10 \) to \( n = 40 \) while keeping the same standard deviation. What happens to the standard error?
Since \( \text{SE} = \frac{s}{\sqrt{n}} \), increasing \( n \) from 10 to 40 means \( \sqrt{n} \) goes from \( \sqrt{10} \approx 3.16 \) to \( \sqrt{40} \approx 6.32 \). That is exactly double, so the SE is halved.
Question 5 of 5
Which graph type is most appropriate for comparing the average reaction times of three different age groups?
A bar graph with error bars is ideal for comparing means across discrete categories (the three age groups). Line graphs are for continuous x-axes, scatter plots for two continuous variables, and pie charts for proportions of a whole.