intro to stats as concept + experimental design

statistics

making meaning out of numbers using logic, math, probability, and research design

Descriptive Statistics

a number (or figure) that conveys/summarizes a particular characteristic of a set of data which is meant to summarize a set of data

Inferential Staistics

Method that uses sample evidence and probability of an actual data set to reach conclusions about unmeasurable populations

want to generalize sample statistics to make claims about a population

population

all the scores of some specified group

typically statistics area of interest

an entire interest group

almost never able to measure entire population

sample

measurement of a subset / group of a population

parameter

some numerical or nominal characteristic of a population

constant/unchanging and cannot be computed

statistic

some numerical or nominal characteristic of a sample

variables

something that exists in more than one amount or form

many ways to define a variable

quantitative variable

degree/amount of the thing being measured

scored indicate different amount

continuous variable

form of quantitative variable

scares can be any value or intermediate value over range (ex, time, height)

specific values, whole numbers

discrete variable

form of quantitative variable

intermediate values between scores do not exits, are not possible, and are not meaningful (ex, amount of siblings, position in race, school name)

full range of values, whole numbers or decimals

operational definition

how we chose to manipulate/measure the variables of interest in a given study

nominal

numbers serve only as labels and do not indicate a quantitative relationship (ex, number on jersey, assign number to universities named)

category or none

no inherent order, just naming, no greater value than another

if assign number, we do not impose value but instead assign “name”

ordinal

characteristic of nominal as well as indicator of a greater or lesser position (ex, position in race)

ordered rankings but not necessarily evenly spaces

greater than vs less than

interval

nominal and ordinal characteristics as well as intervals between numbers are equal

equal differences between numbers represent equal difference between things measured (ex, temperature — 10 degree difference anywhere is still a ten degree difference)

ordered and numbered but don’t know gaps

ratio

nominal, ordinal, and interval as well as a true zero point, meaning none of the thing is present (ex, height, weight)

exists an absolute zero and is not arbitrary

scale

refers to either/both interval or ratio

Scientific method

use a methodological approach to answer questions

how we measure

question formulation —> background reserach —> tentative hypothesis —> test hypothesis —> analyze data —> ask again

Non-experimental studies

correlational

surveys, observations, case studies

no manipulation of variable, no random assignment, simply observing and discovering associations (NOT CAUSE)

correlation =/

causation

correlation CAN ESTABLISH causation but NOT simply by measuring 2 variable

experimental design

at least 1 variable is manipulated by design and at least 1 variable is being measured

purpose: to see cause and effect relationships

independent variable

the variable being manipulated by the experiment

what experimenter will think will cause change

dependent variable

the variable the experimenter will hope to see change

at least 1 DV is measured

control

in repeated events, everything about the experiment should be the same so we make sure the change being made is by the DV

control variable

a base variable with no change to see if the IV is actually working

random assignment

individuals / subjects in the experiment are randomly assigned to their groups

reduces confounding variables or different characteristics to interfere and explain results

random sampling

pick individuals from a population at random

every person in the population should have an equal chance of being selected

extraneous variable

another variable that man effect the DV

w/s EVs, cannot conclude what variable cause the change in the DV

confounding variable

EV that varies systemically with the IV

between group design

different people experience each condition and comparison are made across the two groups

repeated-measure / within group design

same participant experiences all conditions (IV, different IVs, and control) and comparisons are made across the groups but using the different data from the same individual

dont have to worry about random assignment or confounding variable because people do both and eliminate their own extraneous variables that could cause a diff change on the DV

quasi-experiment

based on peoples life choices and compare their outcomes

quasi IV because people already do the IV

ex; test effect of meditation but get people who already meditate