Descriptive Statistics & Uncertainty Analysis
Mathematical Problem Formulation
Statistical Metrics in Experimental Physics:
Given $N$ independent measurements $x_1, x_2, \dots, x_N$ of a physical observable:
1. Sample Mean ($\bar{x}$):
$$\bar{x} = \frac{1}{N} \sum_{i=1}^N x_i$$
2. Sample Standard Deviation ($s_x$ with Bessel's Correction):
$$s_x = \sqrt{\frac{1}{N - 1} \sum_{i=1}^N (x_i - \bar{x})^2}$$
3. Standard Error of the Mean ($\mathrm{SEM}$):
$$\mathrm{SEM} = \frac{s_x}{\sqrt{N}}$$
4. Outlier Rejection via Tukey's Interquartile Range (IQR):
Let $Q_1$ and $Q_3$ represent the 25th and 75th percentiles. A measurement is classified as a gross experimental error (blunder) if it falls outside the Tukey fences:
$$x_{\text{outlier}} \notin [Q_1 - 1.5 \cdot \mathrm{IQR},\; Q_3 + 1.5 \cdot \mathrm{IQR}]$$
5. Error Propagation to Derived Quantities:
For a simple pendulum of length $L$ and period $T$, $g = \frac{4\pi^2 L}{T^2}$. The propagated uncertainty is:
$$\delta g = g \cdot 2 \frac{\mathrm{SEM}_T}{\bar{T}}$$
Theoretical Background & Explanation
1. Precision vs Uncertainty in Experimental Physics:
In scientific measurement, stating a result as a single number without an uncertainty interval is meaningless. Pandas provides specialized functions for statistical reduction:
.mean(): Best estimate of the true physical parameter..std(ddof=1): Sample standard deviation reflecting experimental spread (uses $N-1$ degrees of freedom by default)..sem(): Standard error of the mean, quantifying the uncertainty in the mean estimate.
2. Automated Outlier Rejection:
Student laboratory sessions often suffer from blunders (e.g. premature stopwatch clicks or electrical glitches). Tukey's IQR rule provides a principled, objective method to prune outliers without subjective investigator bias.
3. Error Propagation:
Combining the calculated $\mathrm{SEM}$ with analytical error propagation yields the final experimental report formatted according to international metrology standards: $g = (\bar{g} \pm \delta g)\,\text{m/s}^2$.