Sampling, Frequencies, and Bivariate Correlational Research

Raw Scores:

data that have not yet been transformed or analyzed 

Frequency table:

shows the pattern of the data by indicating how many participants had each score

Outlier:

an extreme score that is either very high or very low in comparison to the rest of the scores. Can lead to false conclusions about results

Grouped Frequency Table:

aka grouped distribution, a way of organizing a large set of data into more manageable groups (tally marks)

Histograms:

a bar graph that depicts only one variable, usually based on scale data, with values on x-axis and frequencies on y-axis

Normal distribution:

bell shaped, symmetrical unimodal curve

Skewed distribution:

one of the tails is pulled away from the center

positive skew

tail extends to right, median and mean are higher than mode

Negative skew:

tail extends to left, median and mean are lower than mode.

Ceiling effect:

when a considerable percentage score the best or maximum possible score

Floor effect:

when a considerable percentage of participants obtain the worst or minimum available score

Dot Plot:

simple form of data visualization that consists of data plotted as dots on x-axis 

Sampling:

process of selecting a limited number of units from a larger set. Selection of respondents, observations, cases for inclusion, surveys, interviews, etc

Population :

entire group you want to draw conclusions from (bag of chips)

Sample :

specific group data is collected from. Generalizable (one chip). Just because it comes from a population doesn't mean its generalizable to population

Census :

systematically acquiring, recording, and calculation population information.

Biased sample:

some members of the population have a much higher probability over other of being included in the study

Unbiased sample:

all members of the population have an equal change of being included in the sample

Convenience sampling:

sample of people who are easy to contact and are readily available to participate (friends, relatives, colleagues, people on street)

Self-Selection Bias :

  participants have control over their participation. People with strong feelings are more likely to participate, and those less passionate will likely not volunteer. 

Undercoverage (exclusion bias/attrition bias) :

various potential biases that can result from post-randomization exclusion of patients from a trial and subsequent analyses.

Nonresponse Bias :

some of the respondents selected fail to respond, cant be contacted, or decline to participate

Random sampling (probability sampling) :

sampling technique where each sample has an equal probability of being chosen

Simple Random Sampling

sample is randomly selected. most basic form of random sampling

Systematic sampling:

researchers select members of a population at a regular interval (ex: selecting every 5th person in line to ask questions)

Stratified sampling:

researches divide subjects into subgroups called strata baked on characteristics they share. (ex race, gender, education). Then each subgroup is randomly sampled using another probability sampling method.

Cluster sampling:

divide a population into clusters, such as districts or schools etc, and then randomly select some of these clusters to be the sample.

Quota sampling:

predetermined number of proportion of units. Create a convenience sample with individuals that represent a population.

Snowball Sampling:

best for cases with hard-to-reach participants. Existing participants are asked to recommend others for the study

Association Claims:

argues that one variable is likely to be associated with a particular level of another

Correlation:

defined as a relation between things or phenomena or things that tend to vary, be associated, or occur together in a way not expected by chance alone

Bivariate correlation:

a technique that determines the existence of a relationship between two variables. (if X changes by how much, then Y will change by how much) 

Scatter plot:

uses dots to represent values for two different numeric variables. Shows relationships between X and Y

correlation coefficient (r)

measures a linear correlation. it is a number between -1 and 1 that measures the strength and direction of the relationship between two variables.

Categorical variables:

represent types of data which may be divided into groups (race,sex,age,education)

Construct validity:

concerns the extent to which your test or measure accurately assesses what its supposed to 

Effect Size:

a value that measures the strength between two variables and how meaningful the difference is.

How do you calculate effect size

Effect size = [mean₂-mean₁] 

The confidence interval (CI) :

the range of values that you expect your estimate to fall between a certain percentage of the time if you run your experiment again or resample in the same way. 

Replication:

the process of repeating research to determine the extent to which findings generalize across time and situations

Three conditions to establish internal validity

covariation, temporal precedence, elimination of alternative explanation

Longitudinal Study:

a type of correlational study where researchers observe and collect data on a  number of variables without influencing them, over a period of time

Retrospective study:

uses existing data that was previously recorded for reasons other than research

Prospective study:

researcher chooses a group of subjects and follows them over time, collecting real time data 

Cross-sectional correlations:

study looks at data from a single point in time

Autocorrelations:

refer to the degree of correlation of the same variables between two successive time intervals 

Cross-Lag Correlations:

show whether the earlier measure of one variable is associated with the later measure of the other variable

A mediator:

a variable that explains the relationship between two other variables

A moderator:

a variable that affects the direction or strength of the relationship between two other variables 

Multivariate design:

involves more than two measured variables