Notes on Expectations, Standard Deviation, and Standard Error of the Mean

Scope of video: covers three core ideas (expectations/mean, standard deviation, standard error of the mean), how to graph them, and how to construct data tables for the class.

Definition: The expectation (mean) is a measure of central tendency and represents the long-run average if the experiment were repeated many times.
Population vs. sample:
- Population mean: \mu = E[X] = \sum{i} xi \;pi for discrete variables, or \mu = E[X] = \int{-\infty}^{\infty} x \; f(x) \, dx for continuous variables.
- Sample mean (estimate of the population mean): \bar{x} = \frac{1}{n} \sum{i=1}^n xi
Properties (conceptual): mean is the point around which observations are centered; sums of deviations from the mean balance to zero (sum of (x_i - \bar{x}) = 0).
Role in data tables/graphs: the mean is often plotted or reported as the central value for a group or condition; used to compare groups when combined with an index of variability.
Relationship to the law of large numbers (intuition): as sample size grows, the sample mean converges to the population mean.

Common visual options:
- Bar graphs or column charts with the mean as the central value for each group.
- Line graphs showing mean across time or ordered categories.
- Overlay of individual data points (dot plots) to show dispersion around the mean.
Error bars on means: indicate uncertainty or variability around the mean (often tied to standard error of the mean or standard deviation).
Considerations:
- Choose metric (mean) appropriate to the scale of measurement and distribution shape.
- Ensure clear labeling, units, and legend if multiple groups are shown.

Purpose: measures how dispersed the data are around the mean; a summary of variability.
Population standard deviation: \sigma = \sqrt{\frac{1}{N} \sum{i=1}^N (xi - \mu)^2}
Sample standard deviation (unbiased estimator of the population SD): s = \sqrt{\frac{1}{n-1} \sum{i=1}^n (xi - \bar{x})^2}
Interpretation:
- Smaller values: observations cluster near the mean.
- Larger values: greater spread of data.
Units: same as the data (since you take a square root of squared deviations).
Relationship to the variance: \mathrm{Var}(X) = \sigma^2 (population) or s^2 = \frac{1}{n-1} \sum (x_i - \bar{x})^2 (sample), with "unbiased" connotation for the sample SD when estimating the population SD.

Purpose: measures the precision with which the sample mean estimates the population mean; reflects sampling variability of the mean, not the spread of individual observations.
Formula (population): SE_{\bar{X}} = \frac{\sigma}{\sqrt{n}}
Estimator when population SD unknown: replace (\sigma) with the sample SD (s): SE_{\bar{X}} \approx \frac{s}{\sqrt{n}}
Interpretation:
- Smaller SEM implies more precise estimate of the population mean from the sample.
- SEM decreases as sample size n increases, with the rate proportional to (1/\sqrt{n}).
Relationship to confidence intervals:
- Approximate 95% CI for the mean (assuming normality or large n): \bar{x} \pm z{0.975} \cdot SE{\bar{X}} where a z-value is used for large samples.
- If population variance unknown and sample size is small: use the t-distribution: \bar{x} \pm t{(n-1, 0.975)} \cdot SE{\bar{X}}
Graphical use: error bars on mean plots commonly show ±1 SEM or ±1.96 SE for approximate 95% CI, but notation must be clear to avoid misinterpretation.

Key columns to include when presenting data:
- Group/Condition identifier
- n (sample size per group)
- Mean (\bar{x})
- Standard deviation (s) or (\sigma) (population, if known)
- Standard error of the mean (SE_{\bar{X}})
Formatting guidelines:
- Clearly label units and scales.
- Report numbers to an appropriate level of precision (e.g., two decimals for means, two decimals for SE/SD).
- Include a note on whether SD is population or sample; specify if SEM is based on sample SD.
Example table structure:
- Group | n | (\bar{x}) | s | SE(\bar{X}) | [optional] 95% CI
Example data snippet (illustrative):
- Group A: n = 5, data = {2, 4, 5, 7, 9}
- Mean: \bar{x} = \frac{1}{5} (2+4+5+7+9) = 5.4
- Differences: (-3.4, -1.4, -0.4, 1.6, 3.6)
- Squared deviations: (11.56, 1.96, 0.16, 2.56, 12.96)
- Sum = 29.2
- Sample variance: s^2 = \frac{29.2}{n-1} = \frac{29.2}{4} = 7.3
- Sample SD: s = \sqrt{7.3} \approx 2.702
- SEM: SE_{\bar{X}} = \frac{s}{\sqrt{n}} = \frac{2.702}{\sqrt{5}} \approx 1.208
- Approx 95% CI (t with df = 4, ~2.776): 5.4 \pm 2.776 \cdot 1.208 \approx (2.05, 8.75)

Relationship to sampling theory: SEM and SD reflect different aspects of variability—data spread vs. precision of the mean estimate.
Practical implications:
- Misinterpreting SEM as data spread can mislead readers about variability within the data.
- Reporting both mean and SD helps convey central tendency and dispersion; SEM helps understand the reliability of the mean estimate across samples.
Ethical/statistical diligence:
- Use appropriate error measures for the context (e.g., SD for variability, SEM for precision of the mean).
- When presenting, be transparent about which measure is used and the sample size, so conclusions are not overstated.

Population mean: \mu = E[X] = \sum{i} xi pi (discrete) or \mu = E[X] = \int{-\infty}^{\infty} x f(x) \, dx (continuous)
Sample mean: \bar{x} = \frac{1}{n} \sum{i=1}^n xi
Population SD: \sigma = \sqrt{\frac{1}{N} \sum{i=1}^N (xi - \mu)^2}
Sample SD: s = \sqrt{\frac{1}{n-1} \sum{i=1}^n (xi - \bar{x})^2}
Standard error of the mean: SE_{\bar{X}} = \frac{\sigma}{\sqrt{n}} \approx \frac{s}{\sqrt{n}}
Confidence interval for the mean (large n): \bar{x} \pm z{0.975} \cdot SE{\bar{X}}
Confidence interval for the mean (small n, t): \bar{x} \pm t{(n-1, 0.975)} \cdot SE{\bar{X}}

Mean (expectation) summarizes the central tendency of a distribution; it is a population parameter (μ) or a sample statistic (x̄).
Standard deviation (SD) describes how spread out individual observations are around the mean; it is σ for population and s for sample.
Standard error of the mean (SEM) describes the precision of the sample mean as an estimate of the population mean; SEM decreases with larger n and is computed as σ/√n (or s/√n as an estimator).
Graphs of means often use error bars to convey variability or precision; choose SD vs SEM carefully and state what the error bars represent.
Data tables should report n, mean, SD (or σ), and SEM with clear labels, units, and rounding rules; provide a transparent example to illustrate calculation steps.
Always connect these concepts to their practical interpretation and be mindful of how conclusions may be affected by sample size and chosen error measures.