Measurement and Statistics in Exercise Science

Flashcards covering key vocabulary and concepts from the 'Measurement and Statistics' lecture in exercise science, including definitions of terms, scales of measurement, validity, reliability, and various statistical tests.

Measurement (in exercise science)

Fundamental to professional practice and research in exercise science, ensuring credibility and guiding exercise prescriptions.

Test

An instrument or tool used to take measurements and gather data.

Measurement

The act of assessment to collect numerical information by assigning numbers to observations according to rules.

Evaluation

A value judgment placed on the measurement, involving interpretation of whatever has been measured.

Statistic

A number (datum) or numbers (data) calculated from measured data.

Statistics

Techniques that deal with the collection, organization, analysis, description, interpretation, and presentation of information stated numerically.

Nominal Scale

Numbers that represent names, forming mutually exclusive categories with no meaningful order (e.g., sex, race).

Ordinal Scale

Numbers with a characteristic of order (higher/lower) but no common unit of measurement between them, so data cannot be averaged meaningfully (e.g., ranking of sport teams).

Interval Scale

Numbers with a meaningful order and a common unit of measurement (equal distance between scores), but an arbitrary zero point (e.g., temperature, IQ scores).

Ratio Scale

Numbers with a common unit of measurement (equal distance) between scores and an absolute zero point, indicating a true lack of the measured attribute (e.g., height, weight, heart rate).

Norm-referenced Standard

Evaluates measurements by comparing them to the performance of others (norms) (e.g., 70th percentile on a vertical jump).

Criterion-referenced Standard

Evaluates measurements by comparing them to a predetermined standard or criterion (e.g., performing 7 or more push-ups to reach a Healthy Fitness Zone).

Validity

The extent to which inferences made from specific measures are appropriate, meaning the test measures what it is supposed to measure.

Content Validity

Evidence based on disciplinary expert judgment that test items represent all important content areas the test claims to measure (e.g., written knowledge tests in exercise science organizations).

Logical Validity

Evidence demonstrated by the extent to which a test is judged to measure the most important components of skill necessary to perform a motor task adequately (e.g., assessing soccer dribbling skill).

Criterion-Related Validity

Techniques to demonstrate the correlation of a test of a construct with a criterion measure of that construct (e.g., using VO2max as a criterion for aerobic capacity).

Reliability

The consistency or repeatability of a measurement.

Intraclass Correlation Coefficient

A statistic used to measure reliability, with professionals generally desiring a value of 0.8 or higher.

Classical Test Theory (X = T + e)

A framework where X is the measured variable, T is the true score, and e is measurement error. The goal is to minimize 'e'.

Measurement Error Minimization

Achieved by assuring instrument calibration, trained testers, standardized procedures, and prepared participants.

Test-Retest Reliability

Assessed by testing participants with two or more trials or on two or more occasions to check consistency over time.

Internal Consistency Reliability

Calculated from a single administration of a test with multiple items (e.g., a multiple-choice test).

Equivalence Reliability (Parallel Forms Reliability)

Estimated by administering two forms of a test designed to measure the same construct to the same people and correlating their scores.

Criterion-Referenced Framework (for tests)

Used to make categorical decisions, such as whether a person passed or failed a test or met a standard.

Descriptive Statistics

Techniques used to organize or summarize a set of measurements (e.g., central tendency, variability).

Inferential Statistics

Techniques used to make inferences from a sample about the larger population it represents.

Central Tendency

A measure that best represents typical or central scores of data.

Mean (Arithmetic Mean)

The sum of all scores divided by the number of scores, typically the best measure of a typical score if data are normally distributed.

Median

The actual middle value of a data set, with 50% of scores falling above and below it.

Mode

The most frequently occurring score in a distribution, useful in certain situations though less common in exercise science.

Standard Deviation (S)

Describes the variability in a data set in the same measurement units as the data set, conceptually understood as the average deviation of each score from the mean.

Standard Score

Scores that have been standardized to a constant mean and standard deviation, converting raw scores into a standard unit of measurement.

Z-score

A type of standard score that indicates how many standard deviation units a score is from the mean; calculated as (X - Mean) / Standard Deviation.

Normal Curve (Bell-shaped curve)

A symmetrical distribution where the mean, median, and mode are at the center (Z-score of zero), with 50% of scores above and 50% below this point.

Percentile Rank

The percentage of scores that fall below any given point on the normal curve; for a Z-score of zero, the percentile rank is 50.

Correlation Coefficient

A statistic that describes the strength and direction of a linear association between two variables, ranging from -1.00 to +1.00.

Pearson Product-Moment Correlation Coefficient

A specific type of correlation coefficient used to describe the linear association between two variables.

Coefficient of Determination (r²)

The square of the correlation coefficient, representing the proportion of variance in one variable that can be predicted from the other.

Positive Correlation

A relationship where as the value of one variable increases, the value of the other variable tends to increase.

Negative Correlation

A relationship where as the value of one variable increases, the value of the other variable tends to decrease.

Correlation vs. Causation

A correlation indicates an association between variables but does not imply that one variable causes the other.

Regression Analysis

A statistical method used to develop equations that allow prediction of one variable from one or more other variables.

Criterion Variable (Dependent Variable)

The variable that a regression analysis tries to predict, often difficult or expensive to measure directly.

Predictor Variables (Independent Variables)

The variables used in a regression analysis to make predictions about the criterion variable.

Simple Linear Regression Equation

A regression equation that predicts a criterion variable from one predictor variable.

Multiple Regression Equation

A regression equation that includes more than one predictor variable to predict a criterion variable.

Effect Size

A quantitative measure of the magnitude of a phenomenon, such as the size of the relationship between variables or the difference between groups.

Cohen’s Delta (Effect Size Calculation)

A common measure of effect size, calculated as the difference between two means divided by the standard deviation.

t-test

A statistical test used to test a null hypothesis stating no difference exists between two means.

Independent Groups t-test

Used when the people in one group are independent of the people in the other group (e.g., different participants randomly assigned to two conditions).

Dependent Groups t-test

Used when the people in one group are related to the people in the other group (e.g., comparing pretest and posttest means from the same individuals).

Analysis of Variance (ANOVA)

A statistical test used in situations where more than two means are involved in a study.

One-way ANOVA

An ANOVA with only one predictor (independent) variable, which may have two or more groups or levels (e.g., comparing physical activity levels across three student/professor groups).

Post Hoc Tests (in ANOVA)

Statistical tests used after a significant ANOVA result to determine exactly which specific means are different from each other.

Factorial ANOVA

An ANOVA that studies the effects of more than one categorical independent variable on one continuous outcome variable; also examines interactions between independent variables.

Interaction (in Factorial ANOVA)

Occurs when the effect of one independent variable on the outcome depends on the level of another independent variable, indicating complex effects.

Repeated Measures ANOVA

An ANOVA used when participants are measured on the same outcome variable on several occasions over time (e.g., measuring on-task behavior on different days).