Statistics Quiz 1

Why should you take a statistics class?

The world is increasingly data driven, statistical know-how is marketable, builds critical thinking skills, builds statistical literacy

The scientific method

is the process by which scientists attempt to reign in biases that emphasizes use of empirical research and data-based conclusions; requires replication

Steps of the scientific method

Curiosity, develop hypothesis, test the hypothesis, analyze data and draw conclusions, report the results

What are some reasons that the scientific method can lead to faulty conclusions?

Conclusions are sometimes drawn too quickly, findings can be population specific, conclusions limited to study design and statistical analysis

Data

refers to sets of values (typically numerical) we assign to a person or thing we are interested in studying

Three types of methodology

self-report, authority, or archival

Operationalization

the process by which we take a meaningful but somewhat vague concept and turn it into a precise measurement`

Theoretical construct

the thing you are trying to take a measurement of, can not be directly observed and is often vague

Measure

the method or tool used to make observations

Variable

what we end up with when we apply our measure to something in the world, the actual data we end up with

4 Scales of measurement

nominal, ordinal, interval, ratio

Nominal

there is no relationship between the categories (ex. Gender of marathon runners)

Ordinal

categories in a specific order (ex. Positions in a race)

Interval

the differences between values/numbers are interpretable and meaningful (ex. Temperature)

Ratio

have a true zero, you can multiply and divide

Types of variable

continuous and discrete

Continuous

can be any number within a particular range (decimals)

Discrete

no in between values (no decimals)

Likert scales

1-5 or very unsatisfied, a little unsatisfied, etc. (discrete)

Reliability

how precisely you are measuring something, consistent and repeatable

Measuring reliability

test-retest, inter-rater, parallel forms, internal consistency

Test-retest reliability

refers to consistency over time

Inter-rater reliability

refers to consistency across people

Parallel forms reliability

refers to consistency across theoretically-equivalent measurements

Internal consistency

refers to consistency across items in a measure

Validity

how accurately you are measuring something, are you measuring what you think you are measuring

Types of validity

internal, external, construct, face, ecological

Most important types of validity

external and internal

Internal validity

whether or not you can draw the correct conclusions about the causal relationships between variables

External validity

the generalizability or applicability of findings

Generalizability

how likely you are to see the same results from your study in the “real world”

Construct validity

whether you are measuring something accurately

Face validity

whether or not a measure looks like it is doing what it is supposed to do

Ecological validity

whether or not the experimental settings accurately reproduces the real world settings

Threats to validity

confounds, artifacts, history effects, non-response bias, regression to the mean, experimenter bias

Confounds

additional, unmeasured variables that influence your results (threaten internal validity; bigger problem for non-experimental study designs)

Artifacts

results that only occur in the special situation of your study (threaten external validity; bigger problem in experimental study designs)

History effects

specific events that occur during the study that influence the outcome

Non-response bias

participants may not respond to research (avoiding difficult topics is a problem)

Regression to the mean

usually extreme values are followed by average ones

Experimenter bias

unintended influence from an experimenter

Methodology- methods of collecting data

experimental designs, observational, archival research, correlational studies, survey methods

Experimental designs

the researcher controls as many aspects of the study as they can, can infer cause and effect

Observational

systematically observing and measuring behavior; frequency, duration, or interval method; naturalistic observations in a natural setting

Archival research

the use of pre-existing data

Correlational studies

exploring the relationship between two variables

Survey methods

involve the recruitment of large sample sizes via surveys

Descriptive statistics

summarizes the data in compact, easily understandable ways

Observations

each value within a variable drawn from our participants

Distribution

the collection of observations for a single variable

Measures of central tendency

mean, median, and mode

Mean

average (interval or ratio)

Median

middle value (ordinal; interval or ratio)

Mode

most frequent value (nominal)

Measures of variability

range, interquartile range, mean absolute deviation, variance, standard deviation

Range

biggest value minus the smallest value (worst measure of variability)

Interquartile range

difference taken between the 25^th and 75^th percentile

Mean absolute deviation

average distance between each data point and the mean

Variance

average of squared deviations from the mean (observations – 1)

Standard deviation

square root of variance; 68% of data within 1SD, 95% within 2SD, and 99.7% within 3SD

Skew

measure of the asymmetry of data; a skew of (+/-) 2 is a sign of asymmetry

Negative skew

majority on the right of the graph

Positive skew

majority on the left of the graph

No skew

equal on both sides

Kurtosis

measure of pointiness of data; normal curve should have kurtosis of 0

Platykurtic

flattened, kurtosis will be negative

Leptokurtic

too pointy, kurtosis will be positive

Mesokurtic

‘just right,’ kurtosis will be 0

Standard scores (z-scores)

number of standard deviations above/below the mean; tell how extreme your score is and allows a comparison between scales

Histograms

use interval or ratio data, bars indicate the frequency of observations, give an indication of skew and kurtosis

Boxplots

use interval or ratio data; shows median, IQR, and range; useful in identifying outliers

This is the first step in the scientific method, which involves posing questions

Curiosity

1.       This is the third step of the scientific method, and involves gathering evidence

a.       Testing the hypothesis

This is the fifth step of the scientific process, and involves sharing data and analysis

a.       Reporting results

1.       This is the assigning of labels, number, or some other well defined descriptions to the things you are interested in

a.       Measurements

1.       This is what you end up with when you apply a measures (it is what you get when you get data)

a.       Variables

1.       In operationalization this is the method or tool used to make your observations (a survey instrument)

a.       Measure

1.       In operationalization, this is something vague but meaningful

a.       Theoretical construct

1.       This variable scale of measure includes ordered categories, like places in a race

a.       Ordinal

1.       This variable scale of measurement involves meaningful numerical differences (no true zero)

a.       Interval

1.       This variable scale of measurement involves simple categories, like gender

a.       Nominal

1.       Involves meaningful differences and a true zero

a.       Ratio

1.       True/false- speed of a car in mph is on the interval scale of measurement

a.       False- it’s ratio

1.       True/false- political affiliation is on the nominal scale of measurement

a.       True

1.       True/false- eye color is only ordinal scale of measurement

a.       False

1.       True/false- a continuous variable is which there can be values within values (decimals)

a.       True

1.       True/false- time in seconds is a discrete variable

a.       False

1.       True/false- place in a race is a discrete variable

a.       True

1.       This refers to how accurately you are measuring something (are you measuring what you think you are measuring

a.       Validity

1.       This refers to how precise your measure is (is it consistent and repeatable)

a.       Reliability

This form of reliability refers to consistency across people

a.       Inter-rater reliability

1.       This form of reliability refers to reliability across items in a scale/measure

a.       Internal consistency

1.       This form of reliability refers to consistency over time or repeated measurements

a.       Test-retest

1.       Which answer below refers to parallel forms reliability

a.       Comparison between two measures that are measuring the same thing

1.       Which answer below refers to test-retest reliability

a.       Comparison between test 1 and test 2

1.       In this research design the researcher uses pre-existing data to explore their topic of interest

a.       Archival research

1.       In this research design the researcher tries to control all aspects of the student and introduces a manipulation (cause and effect)

a.       Experimental

1.       The relationship between two variables are being explored

a.       Correlational

1.       True/false-  you can infer cause and effect from correlational

a.       False

1.       This form of validity refers to the extent to which you are able to draw cause and effect conclusions about the results

a.       Internal validity