Glossary
Bar graphs
A visual representation using rectangular bars of varying heights or lengths to compare discrete categories of data.
Example:
A bar graph effectively shows the number of students who prefer different subjects, with separate bars for Math, Science, and English.
Confounding variables
Hidden or unmeasured factors that affect both the independent and dependent variables, making it difficult to establish a true cause-and-effect relationship.
Example:
In a study linking coffee consumption to heart disease, stress levels could be a confounding variable because stress might also increase heart disease risk and be correlated with coffee intake.
Control groups
A group in an experiment that does not receive the treatment or intervention being tested, serving as a baseline for comparison with the experimental group.
Example:
In a study testing a new teaching method, the students taught with the traditional method form the control group.
Controls
Elements within an experiment designed to minimize the influence of confounding factors and ensure that observed changes are due to the independent variable.
Example:
In a drug trial, giving a placebo to one group serves as a control to account for the psychological effect of receiving treatment.
Convenience sampling
A non-probability sampling method where participants are chosen based on their easy accessibility and proximity to the researcher, often leading to bias.
Example:
A student conducting a survey by asking only their friends to participate is using convenience sampling, which may not represent the broader student body.
Data Validity
The extent to which data accurately measures what it is intended to measure and is free from significant errors or biases.
Example:
Before drawing conclusions, a student checks the data validity of a survey by examining the source, collection methods, and potential for manipulated scales.
Dependent variable
The variable that is measured or observed in an experiment, whose value is expected to change in response to the manipulation of the independent variable.
Example:
In the fertilizer experiment, the height of the plant after a certain period is the dependent variable, as it's expected to change based on the fertilizer amount.
Effect sizes
A quantitative measure of the strength or magnitude of a phenomenon or relationship between variables, indicating the practical significance of a finding.
Example:
While a new teaching method might show a statistically significant improvement, a small effect size would indicate that the improvement in test scores is not very meaningful in practice.
Generalizability
The extent to which the findings of a study can be applied or extended to a larger population or different settings beyond the specific sample studied.
Example:
A study conducted on college students in one city may have limited generalizability to all young adults across the country.
Independent variable
The variable that is intentionally changed or manipulated by the researcher in an experiment to observe its effect on another variable.
Example:
In an experiment testing fertilizer's effect on plant growth, the amount of fertilizer applied is the independent variable.
Internal validity
The degree to which a study accurately establishes a cause-and-effect relationship between the independent and dependent variables, free from confounding factors.
Example:
A well-designed experiment with strong controls and randomization has high internal validity, allowing researchers to confidently say the treatment caused the observed effect.
Line graphs
A graph that uses points connected by lines to show how a quantity changes over a continuous period, typically time.
Example:
To track the growth of a plant over several weeks, a scientist plots a line graph showing its height on consecutive Mondays.
Measurement Errors
Discrepancies between the true value of a measurement and the recorded value, often caused by faulty instruments, inconsistent procedures, or human mistakes.
Example:
Using a ruler with a chipped end to measure plant growth could lead to measurement errors, consistently underestimating the true height.
Pie charts
A circular graph divided into sectors, where each sector represents a proportion of the whole, illustrating parts of a total.
Example:
A pie chart displays the breakdown of a student's daily screen time, showing what percentage is spent on studying, gaming, and social media.
Random sampling
A sampling technique where every individual in the population has an equal chance of being selected, ensuring the sample is representative and minimizing bias.
Example:
To survey student opinions on a new school policy, drawing names from a hat to select participants is an example of random sampling.
Randomization
The process of assigning participants to different groups or conditions in an experiment purely by chance, helping to distribute confounding variables evenly.
Example:
Using a coin flip to decide which participants receive the new treatment versus the placebo ensures randomization and reduces bias.
Researcher bias
Occurs when the researcher's expectations or beliefs unintentionally influence the study's design, data collection, or interpretation of results.
Example:
If a scientist strongly believes a new drug will work, they might unconsciously interpret ambiguous results in favor of the drug, demonstrating researcher bias.
Response bias
A type of bias where participants provide inaccurate or untruthful answers, often due to social desirability or misunderstanding.
Example:
In a survey about healthy eating habits, participants might overreport their vegetable intake due to response bias, wanting to appear healthier.
Sample size
The number of individuals or observations included in a study or experiment, directly impacting the reliability and precision of the results.
Example:
A study with a sample size of 10,000 participants is generally more reliable than one with only 50, as larger samples reduce random error.
Scatterplots
A graph that displays the relationship between two numerical variables for a set of data, with each point representing a pair of values.
Example:
A researcher creates a scatterplot to see if there's a correlation between hours studied and test scores, with each dot representing one student's data.
Selection bias
Occurs when the sample used in a study is not representative of the larger population, leading to skewed results.
Example:
A survey about favorite school lunches conducted only among students in the cafeteria during pizza day might suffer from selection bias, as it excludes those who bring lunch or prefer other options.
Significance (p-values)
A statistical measure (p-value) indicating the probability of obtaining observed results (or more extreme) if the null hypothesis were true, used to assess if a result is likely due to chance.
Example:
A p-value of 0.01 suggests that there's only a 1% chance the observed results occurred randomly, leading researchers to conclude the findings are statistically significant.
Stratified sampling
A method where the population is divided into distinct subgroups (strata), and then a random sample is drawn from each stratum to ensure representation.
Example:
To survey opinions on a new curriculum, a school might use stratified sampling by randomly selecting a proportional number of students from each grade level (freshman, sophomore, junior, senior).
Tables
An organized display of data using rows and columns, ideal for presenting precise numerical information and facilitating direct comparisons.
Example:
A student uses a table to compare the average daily temperatures in different cities over a week, making it easy to spot the warmest and coolest days.
Type II errors
A statistical error that occurs when a researcher fails to reject a false null hypothesis, meaning they miss a real effect or relationship that actually exists.
Example:
A drug trial might commit a Type II error if it concludes a new medication has no effect, when in reality it does, perhaps due to a small sample size.
Uncertainty (confidence intervals)
A range of values (confidence interval) within which the true population parameter is estimated to lie with a certain level of confidence, reflecting the precision of an estimate.
Example:
A survey reports that 60% of students prefer online learning with a 95% confidence interval of ±3%, meaning the true preference is likely between 57% and 63%.
Well-supported conclusions
Conclusions that logically follow from the presented data, acknowledge study limitations, and have ruled out plausible alternative explanations.
Example:
A research paper presents well-supported conclusions by clearly linking its findings to the data, discussing potential biases, and suggesting areas for future research.