EXPH 2106- Lec. 6 (Scientific Measurement)

Theory

A particular conception or view of a phenomenon

-abstract, general thinking

-a system of rules or principles about how something works

Hypothesis

proposed explanation for a particular fact or event

-what information do you need to answer your question, test your proposal or prediction?

What information do you need to answer your question, test your prediction?

Data

Data

a collection of facts about an object or phenomenon under study

-requires measurement procedures and systematic recordings [facts are always subject to interpretation]

Variables

phenomenon you are trying to measure in some way

Independent variables

factors you are manipulating as the researcher

-the difference between experimental groups

Dependent variables

the variable being tested and measured as the result

-outcome measures

Types of relationships

-Causal

—>indirect

—>direct

-Non-causal

—>confounding variable

Causal Relationship

A relationship where one variable directly influences another; changing one parameter (independent) will lead to a change in outcome measure (dependent)

Direct relationship

as one variable goes up, so does the other

Indirect relationship

as one variable goes up, the other goes down

Confounding variable

an outside factor that relates the independent variable to the dependent variable when they are not causally related to each other directly

What would be a confounding variable in an experiment done to analyze weight loss from energy expenditure through exercise?

caloric intake

Sources of error/bias

-Performer (hawthorne effect)

-Experimenter (rosenthal effect)

-Equipment

Performer Error

subjects behave differently when they know they are being watched/tested

Hawthorne Effect

know they are being watched/tested

-subjects aren’t acting as “normal” people

Experimenter error

self-fulfilling prophecy — researcher expectations influence their interpretation of the results

Rosenthal effect

researcher interprets data to align with hypothesis

Equipment error

calibration, sensitive enough to detect change, human error in applying

-can lead to systematic error in data

Intra-subject Reliability

variations within a given subject’s performance

Intra-Rater Reliability

variations within a given tester

Inter-Rater Reliability

variations between testers

Validity

the extent to which the test measures what the researcher intends to measure

Example of validity

testing balance —> looking at the strength/function of back extensors = NOT VALID

Acquisition

early learning practice trials

-typically graphed as a performance curve of trials within a session

-does NOT necessarily represent a permanent change in capacity for skilled movement

Retention

the persistence of original learning over a period of no practice

-RETEST the dependent variable after some interval has elapsed during which the skill was not performed

-measures LEARNING

Transfer Tests

how effectively one can carry over learning of a skill from one condition to another

-test the dependent variable in a new performance situation or related skills unit

-sometimes called generalization or adaptation

-measures LEARNING

Performance vs. Learning

-Acquisition

-Retention

-Transfer Tests

Accuracy Errors

-Constant Error (CE)

-Variable Error (VE)

-Total Variability (E)

Constant Error

the average magnitude and direction of error relative to the target goal

Variable Error

the average magnitude and direction of error relative to the performance mean

Total Variability

a combination of VE and CE; that measures “overall accuracy” as it encompasses both consistency and success in achieving the target goal