Module 6, Foundations of Quantitative Research Flashcards
Foundational Concepts of Quantitative Research
- sampling
- variables
- hypothesis testing
- correlation and causation
Foundational Concept: Sampling (population and sample)
- studying a sample of the population to provide a quantitative description about that population
population: the total number of possible units or elements that could be included in a study
- the units or elements is typically people (teams, organizations (classrooms), universities)
sample: a subset of the population used to represent the population
- we want to make inferences about a population
- if you are in sample you are in population
- for population, could impose inclusion criteria could be very restricted or quite broad
Random Sampling
- participants are randomly selected from a population
- want sample to represent larger population
◦ make inferences from the
sample about the larger
population - random sampling or random is the equal probability that any individual or unit from the population will be put into the sample or selected
- it is the least likely to happen
◦ in the process of random
sampling you have to know
who the entire population is
(harder with something like
selecting people with type II
diabetes), the people who
decide to consent are
different
Stratified Random Sampling
- population is divided on a characteristic and then randomly sampled
◦ example: how attitudes
toward physical activity
change throughout university - we do this to get equal representation when we want to do statistical analyses for example
- this way there is not an over representation of a particular group (still same issues of random sampling though)
- you have to know who is in each group and consent remains an issue
Systematic Sampling
- example: pick ever 100th person
- have a list of folks and say i am going to pick every one in ten for example
- not as common
Purposive Sampling
- selection is based on specific criteria
◦ every study has inclusion
criteria, this is not specific
criteria (it is very specific
inclusion criteria) - information rich cases
- have a reason why are selecting these cases
- examples: snowball sampling, quota sampling, expert sampling
Convenience Sampling
- selection is based on easy access to participants
- non-probability sampling
- most research is based on convenience samples
Foundational Concept: Variables
variable: a property or characteristic that can take on different values
- a variable has to vary in values
- it is quantitative in nature
- example: look at different V02 max levels, performance measures, also things that are not directly measureable (team coherence)
Independent Variable (IV)
- the variable or “part of the experiment” you are manipulating
◦ the “cause” of the dependent
variable (DV) scores
◦ levels of the independent
variable
◦ based on categories (the
levels of IV I have is based on
the level of groups I have)
◦ control groups or placebo
groups are part of IV
Dependent Variable (DV; outcome variable)
- the “effect” of the independent variable (IV)
- measured by the researcher
- the outcome of the manipulation
Do athletes with patellofemoral syndrome (PFS) experience differences in knee pain if they wear KT tape or a knee brace?
- What is the IV/DV?
- How many Level of the IV?
- What Kind of Relationship?
- the independent variables are the KT or a knee brace (varying) (type of knee support) -> knee pain (DV)
- 2 levels of the IV
- arrow applies casual direction (direction of IV towards DV not other way around)
What type of variables are used in correctional (non-experimental) studies?
predictor and criterion variables
Predictor Variable
- the presumed “cause” in a correlational (non-experimental) study
- measured by the researcher
- if you are not manipulating anything, presumed IV in a non-experimental study
- it is common to have several predictor variables
Criterion Variable
- the presumed “effect” in a correlational (non-experimental) study
- measured by the researcher
- it is not so obvious sometimes which one is predictor or criterion (pay attention to the arrow because even though it is not casual it can still help us)
time post surgery starting ACL rehab (predictor) -> time from surgery to return to play (criterion)
Purpose: “to examine the association of total and specific types of physical activity, including walking or bicycling, exercising, work or occupational activity, home or housework, and leisure time inactivity with the risk of age-related cataract in women and men.”
Participants: “A total of 52 660 participants (23 853 women and 28 807 men) 45 to 83 years of age from the Swedish Mammography Cohort and the Cohort of Swedish Men.”
- what type of study is this? correlational? experimental?
- what are the predictor/IV variable(s)? predictor
- what are the criterion/DV variable(s)? criterion
- who can the findings from this study be generalized to?
- correlational in nature
- walking or bicycling, exercising, work or occupational activity, home or housework, and leisure time and total time physical activity
- risk of age-related cataract in women and men (if you develop it or you do not)
- 45-83 (middle to old-aged people), geographical location, access to healthcare etc.
Extraneous Variables
- variable that might effect the DV that is not the IV
- variable that explains the relationship between the IV and DV but is not apart of the study
- they are alternative explanations for your findings but they are not apart of the study (are not interested in them)
Extraneous Variables may arise from:
1) participant variables
◦ ex. age, gender, sex, physical
fitness, health status
2) situational variables
◦ ex. temperature of testing
room, water availability,
wind, time of day
3) experimenter variables
◦ ex. appearance and conduct
of the researcher,
encouragement
- things that could impact DV but it is not what you are interested in looking at in terms of the IV
- think of what else can impact your DV
Controlling for Extraneous Variables
1) can be controlled through the design of the study
- examples: experimental setting, instructions to participants, experimenter-participant interactions, participant selection criteria, random assignment with a control group
- held constant (through the selection/inclusion criteria - so certain people may not be selected as a result / eliminate different participants extraneous variable)
◦ is preferred method
2) can be controlled through statistical analysis
- ‘control variables’
- requires you to measure the variable
- measure something and try to control it statistical
Random Assignment
random assignment is different than random sampling (how you get people into the study)
- random assignment: we already have the sample, there is an equal probability of people being placed into any group (the groups are equivalent (amongst each group if equivalent not everyone) - equal which helps reduce extraneous variables)
Moderators
- variables that affect the strength of the relationship between IV/predictor & DV/criterion
- usually categorical in nature
exercise -> desire to eat
(age - moderator)
Mediators
- variables that explain the relationship between IV & DV
- accounts for the relationship between the IV/predictor and the DV/criterion
- helps to think of mediator as the mechanism between IV/predictor & DV/criterion
- ‘why?’
- predictor is going to the mediator (predictor and criterion at the same time) and mediator goes to criteria
Note on Variables
- identification of what the IV, DV, extraneous, control, mediator, and moderator variables is
dependent on:
◦ our research question
◦ is specific to the context of
your study - example: age
◦ an be a predictor, control, or
a moderator variable
depending on the research
question - do not need to have moderator or mediator
Hypothesis Testing
the process quantitative research is founded on the assumptions of hypothesis testing
Research Hypotheses
a statement regarding an expected or predicted relationship between variables (has to be very specific)
research hypotheses specify (as precisely as possible) the nature (is there a relationship or is there not between variables) and direction of the relationship between the variables
Research hypotheses can come from a variety of sources:
- identifying a question or issue to be examined
- review and evaluating relevant theories and research (how these variables are actually related to one another)
A. theories helps us define
what variables we might
need to measure and
define relationships
between variables
B. justify the hypothesis (use
theories in previous
research)
A related to B versus A causes B
- correlational design (associated can also be in place of related)
- experimental design
Time post ACL surgery to start rehab -> time to return to play
- is this correlational or experimental?
- what kind of variables are they?
correlational design / predictor and criterion
There is a positive relationship between the time after ACL surgery to start rehabilitation and the time the athlete returns to their sport
There is a relationship between the time after ACL surgery to start rehabilitation and the time the athlete returns to their sport
Which one is directional and non-directional?
- directional hypothesis
- non-directional hypothesis
Athletes with PFS who use KT tape will experience less knee pain than athletes with PFS who use a knee brace
There will be a difference in knee pain between athletes with PFS who use KT tape and those who use knee bracing
Are these experimental or correlational?
Which one is non-directional and which one is directional?
experimental
1. directional
- want to state which group is scoring higher or lower on the outcome variable (instead of saying positive or negative relationship)
- the less part indicates direction
2. non-directional
A group of researchers were interested in seeing if caffeinated underwear would help participants
lose weight. participants were divided into a control group (wore their regular underwear) and an experimental group (wore caffeinated underwear)
1. What is the design of the study?
2. What is the independent and dependent variable?
3. Write a directional research hypothesis
- experimental (manipulating type of underwear’s)
- 2 levels of the IV (type of underwear’s) - lose weight
- those who wear caffeinated underwear will lose more weight than those who wear normal underwear’s OR the participants who use caffeinated underwear will experience more weight less than those who wear normal underwear
Correlation ≠ Causation
- the strength of the association between variables
- direction of the relationship (positive, negative)
- correlations coefficients range (-1 to +1; 0 = no relationship) - represented by lower case ‘r’
◦ if there is no correlation that
coefficient will be zero
◦ only ranges from +1 to -1
(the more you move away
from 0 the stronger the
relationship)
◦ positive correlation
coefficient - positive
relationship
◦ negative correlation
coefficient- negative
relationship
◦ horizontal line for no
correlation
3 Criteria to Establish Cause and Effect
- the cause (IV) must precede the effect (DV) in time
- the cause (IV) and effect (DV) must be correlated with each other (has to be a relationship there)
- the change seen in the effect (DV) cannot be explained by another variable (tricky; extraneous variables - can we try to identify the other alternative explanations for these particular findings - how do you control for these and isolate)
Purposive Sampling: Snowball Sampling
e.g., identify one person with the characteristic of interest, and have that person connect with others in their own network; for example, a woman with breast cancer might ask all the members of her survivorship group at the hospital to participate in a study
Purposive Sampling: Quota Sampling
e.g., identifying a certain number or representation needed for the study and then sampling up to that number; for example, needing 10 women to complete a fitness test
Purposive Sampling: Expert Sampling
e.g., identifying people with known experience and expertise in an area of interest; for example, asking all heart surgeons from a local hospital to complete a survey
