Reason About Data and Draw Conclusions From Them - Chem & Physics

1. Which measure of central tendency is least affected by extreme outliers: mean or median?

   Median. The median, as the middle value, remains stable despite extreme values, unlike the mean which shifts toward outliers.

2. What does a $95\%$ confidence interval for a mean represent in frequentist terms?

   $95\%$ of such intervals would contain the true mean over repetitions. In frequentist statistics, a 95% confidence interval means that if the experiment is repeated many times, 95% of intervals will capture the true population mean.

3. What conclusion is supported when a scatterplot shows a strong negative linear trend?

   As $x$ increases, $y$ tends to decrease (negative correlation). A strong negative linear trend in a scatterplot indicates that the variables move in opposite directions, supporting a negative correlation.

4. Identify the axis transformation used when equal distances represent equal ratios of a variable.

   Logarithmic axis scaling. Logarithmic scaling compresses exponential relationships, making equal ratios appear as equal intervals on the axis.

5. What does it mean if an error bar is labeled as standard deviation rather than SEM?

   It reflects data spread, not uncertainty in the estimated mean. Standard deviation error bars show the variability within the sample data, whereas SEM indicates the uncertainty of the mean estimate.

6. What is the definition of an outlier in a dataset for MCAT-style interpretation?

   A point far from the overall pattern or distribution of the data. Outliers deviate markedly from the main cluster of data points, potentially influencing trends or statistical measures.

7. What does it mean if the correlation coefficient is $r = 0$ for two variables?

   No linear association is present between the variables. A correlation coefficient of $r=0$ implies no linear relationship exists between the variables.

8. What does a $p$-value represent under the null hypothesis?

   Probability of results at least as extreme as observed, assuming $H_0$. The p-value quantifies the evidence against the null hypothesis by calculating the probability of observing the data or more extreme under H0.

9. Calculate the slope if a line passes through $(2,5)$ and $(6,13)$.

   $2$. The slope is calculated as the change in y divided by the change in x between the two points.

10. Find the percent increase when a value changes from $40$ to $50$.

    $25\%$. Percent increase is derived from the relative difference between the new and old values, multiplied by 100%.

11. Compute the fold change when a signal increases from $3$ units to $12$ units.

    $4$-fold. Fold change is the ratio of the final value to the initial value, indicating how many times the signal has multiplied.

12. What is the formula for fold change from baseline $x_0$ to new value $x_1$?

    $\frac{x_1}{x_0}$. Fold change measures the ratio of a new value to the baseline, indicating multiplicative increase or decrease.

13. What conclusion is supported when a scatterplot shows a strong positive linear trend?

    As $x$ increases, $y$ tends to increase (positive correlation). A strong positive linear trend in a scatterplot indicates that the variables move in the same direction, supporting a positive correlation.

14. Which conclusion is supported when two $95\%$ confidence intervals do not overlap?

    The group means likely differ significantly (often $p < 0.05$). Non-overlapping 95% confidence intervals suggest the true means are different, typically indicating statistical significance at p<0.05.

15. What does it imply if a dose-response curve plateaus at high dose?

    A maximum effect is reached (system saturation). A plateau in a dose-response curve occurs when further increases in dose yield no additional response due to receptor or system saturation.

16. Which conclusion is valid if two variables are correlated but no mechanism is established?

    Association only; causation cannot be inferred from correlation alone. Correlation establishes association but requires additional evidence, such as controlled experiments, to infer causation.

17. Choose the correct conclusion if a treatment group mean differs but $p = 0.20$.

    Not statistically significant at $\alpha = 0.05$; fail to reject $H_0$. A p-value greater than the significance level α=0.05 provides insufficient evidence to reject the null hypothesis.

18. Identify the best conclusion if a graph shows an apparent trend but $n$ is very small.

    The trend is tentative; low $n$ limits reliability and generalizability. Small sample sizes reduce statistical power, making observed trends less reliable and harder to generalize to the population.

19. What is the formula for percent change from $x_0$ to $x_1$?

    $\frac{x_1 - x_0}{x_0} \times 100\%$. Percent change quantifies the relative difference from an initial value, expressed as a percentage.

20. What is the slope formula between points $(x_1,y_1)$ and $(x_2,y_2)$ on a graph?

    $\frac{y_2 - y_1}{x_2 - x_1}$. The slope represents the rate of change in $y$ per unit change in $x$ between two points.

21. Identify the correct interpretation if a log-log plot is linear with slope $-2$.

    $y \propto x^{-2}$. A linear log-log plot with slope -2 indicates an inverse square relationship between the original variables.

22. What does a larger standard deviation indicate about a dataset's spread?

    Greater variability; data are more dispersed around the mean. Standard deviation measures the average distance of data points from the mean, with larger values indicating wider dispersion.

23. What is the standard error of the mean (SEM) in terms of $s$ and sample size $n$?

    $\frac{s}{\sqrt{n}}$. SEM estimates the precision of the sample mean as an approximation of the population mean, decreasing with larger sample size.

Study these flashcards