Week 3 Flashcards
a good quantitative design is one that: (4)
- appropriately tests the hypothesis or answers the research questions (design matches the question)
- lacks bias (random selection)
- controls extraneous or confounding variables (any interval that can interfere w the IV effect on the DV)
- has sufficient ability to detect statistically signif. findings (is the sample large enough?)
what is the importance of sampling (2)
- the key to external validity
- considers the degree to which the sample being studied is representative of the population from which it was drawn
describe what plays a role in sampling (4)
- carefully define the population to which you wish to generalize the results (theoretical population)
- define the population to which you have access (accessible population)–> compare and contrast it to the theoretical population
- describe the method used to access the population
- identify a method of selecting & accessing individuals from the population of potential subjects available
what are 2 types of sampling strategies
- probability sampling
- nonprobability sampling
what is probability sampling
- Probability sampling is a sampling technique, in which the subjects of the population get an equal opportunity to be selected as a representative sample.
what is nonprobability sampling? when is it useful?
- involves the intentional selection of certain participants in order to gather information about members of a specific group or people with specific insight into a particular area.
- useful in conditions where you may not have access to an entire population to conduct random sampling or if the researchers are interested in participants who have a certain area of knowledge or expertise
- specifies inclusion and exclusion criteria
what are 4 examples of probability sampling
- random sampling
- stratified random sampling
- cluster sampling
- systematic sampling
what is random sampling
- gold standard
- every potential participant has an equal chance of being selected
what is stratified random sampling
- involves dividing your population into different known subgroups proportionately by categories such as hair color or eye color and then taking a single random sample from each subgroup
what is cluster sampling
- similar to stratified random sampling, in cluster sampling, the researchers divide the total population into subgroups
- rather than selecting members of categorically organized groups, researchers choose entire subgroups of non-organized people to be the participants
what is systematic sampling
- select every __th case from a list
what are 3 types of nonprobability sampling
- convenience sampling
- consecutive sampling
- snowball sampling
what is convenience sampling
- select the most conveniently available people
- includes minimal inclusion/exclusion criteria
ex. go to the wall and ask people
what is snowball sampling
- subjects are asked to recommend other potential subjects
- “word of mouth” recruitment
what is calculated to indicate how many subjects are needed
- sample size estimates –> power analysis
what happens if you have too small of a sample? too large?
- too small = can lead to committing an error in the results
- too large = needless expense, ethical issues,
what are 2 main ways to collect quantitative data
- existing data (hospital records, charts, systematic reviews)
- new data/research
what are 4 types of quantitative data
- self-report
- pt reported outcomes
- direct observation
- biophysiological measures (ex. BP, T, HR, RR)
what are 2 examples of structured self-reports
- interview schedule
- questionnaire
describe the differenve between an interview schedule and questionnaire
- interview = questions pre-specified but asked orally, either face-to-face or telephone
- questionnaire = questions pre-specified but in written form, and independently completed by respondents
what are 2 types of questions in a structured instrument
- close-ended
- open-ended
describe close-ended questions (4)
- fixed-alternative questions (quantitative)
- pre-specified response alternatives
- delivered in the same order w the same response options for every participant
- numerically scored
what are 5 types of closed-ended questions
- dichotomous (2 options, yes/no, fixed responses)
- multiple choice
- rank order questions (ex. rank order of importance)
- forced-choice question (ex. do you prefer tea, hot chocolate, or coffee?)
- rating question (0-10 scale)
define: scale
- a device that assigns a numeric score to people along a continuum
- used to make fine quantitative discriminations among people w different attitudes, perceptions, traits, etc.
what is the likert scale
- consists of several declarative statements (items) expressing viewpoints
- responses are on an agree/disagree continuum
- responses to items are summed to compute a total scale score
what is a visual analog scale
- used to measure subjective experiences (ex. pain, nausea)
- measurements are on a straight line measuring 100 mm
- participants are asked to place an X on the line (or some mark) that reflects where they’re at
what are 3 main biases that can distort measurement
- social desirability bias
- extreme response set bias
- acquiescence response set bias
what is the social desirability bias
- when there is the bias to answer in ways acceptable to societ
what is the extreme response set bias
- a tendency to always answer in the extreme
ex. always select strongly option
what is the acquiescence response set bias
- always agree or always disagree without considering what the question asks
what strategy can be used to counteract the acquiescence response set bias
- counterbalance the - and + response –> forces them to read the Q
what do observational methods focus on
- structred observation of pre-specified behaviors
what is an issue associated w observational methods
- concealment or reactivity –> changing behavior based on the knowledge of observation
what are various methods of recording observations
- senses
- paper/pencil
- video
- audio
- pictures
what are 2 types of biophysiological measures
- in vivo
- in vitro
what does in vivo mean
- measurements directly on or within living organisms
what does in vitro mean
- extracting material and taking it to a lab
what does quantitative measurement involve (5)
- rules for assigning numeric values to the quantity of an atribute
- research instruments should capture the concept in a relevant and truthful way
- objective & precise info –> facilitates statistical analysis
- ALL instruments have some amt of error
- challenging when measuring subjective concepts or psychosocial concepts
obtained score =
obtained score = true score +- error
what are 3 main factors affecting data quality in quantitative research
- procedures used to collect the data (properly trained? told participants to read carefully?)
- circumstances under which data were gathered (ex. may not do well if in cold room)
- adequacy of instruments or scales used to measure constructs
define reliability of an instrument
- an instrument that consistently and accurately measures the concept or construct of interest
- extent to which scores are free from measurement error (accuracy) and consistency
define accuracy
- extent to which the instrument is free from measurement error
define consistency
- extent to which the instrument can consistently (repeatedly) measure the concept or construct
measures should have reliability cofficients of at least ___ or better
what often indicates this?
- at least 0.70 or better
- often indicated thru Cronbach’s alpha
define: validity
- an instrument that actually measures what it is supposed to measure
ex. if you choose a valid measure to measure depression, it should measure depression and not anxiety
what are 4 types of validity of instruments
- face validity
- content validity
- criterion validity
- construct validity
describe face validity
- whether the instrument looks like it is measuring the target construct
describe content validity
- the extent to which the instrument’s content adequately captures the construct
- focuses on how well each question taps into the specific construct in question.
- asks the literature or content expert
describe criterion validityq
- the extent to which the scores on a measure are a good reflection of a “gold standard”
- compare to gold standard
describe construct validity
- the degree to which evidence about a measure’s scores in relation to which other variables supports the inference that the construct has been well represented
- does it reflect what theory says it should look like?
what are the purposes of statistical analysis in quantitative research (4)
- to describe the data (ex. sample characteristics)
- to test hypotheses
- to provide evidence regarding measurement properties of quantified variables
- test the reliability & validity of measurement tools
what are 4 lvls of measurement
- nominal
- ordinal
- interval
- ratio
describe the nominal lvl of measurement
- lowest lvl
- involves using numbers simply to categorize attributes, dont rlly mean anything
ex. numbers to categorize gender, color, clothing, etc.
describe the ordinal lvl of measurement
- ranks people on an attribute
- do not know the distance between each rank
ex. mild, moderate, severe, ADLs range from total dependence to independence
describe the interval lvl of measurement
- ranks people on an attribute and specifies the distance between them
ex. measurement of intelligence, temperature in celsius
describe the ratio lvl of measurement
- highest lvl
- has an absolute zero
- provides info about the absolute magnitude of the attribute
ex. cell count, cases of H1N1 in MB
what are 2 types of statistic analysis
- descriptive statistics
- inferential statistics
what is descriptive statistics
- used to describe and summarize data
what is a parameter
- descriptor for a population
what are inferential statistic
- used to make inferences about the population based on sample data
- helps test the research Q
what is the central tendency r/t descriptive statistics
- index of “typicalness” of a set of scores that comes from center of distribution
- includes mode, median, mean
what is mode
- the most frequently occurring score in a distribution
- useful mainly as gross descriptor, especially of nominal measures
what is median
- the point in a distribution above which & below which 50% of cases fall
- useful mainly as descriptor of typical value when distribution is skewed)
ex. household income - ensure they are ordered in their natural order*
what is mean
- equals the sum of all scores divided by the total number of scores
- most stable and widely used indicator of central tendency
define variability
- the degree to which scores in a distribution are spread out or dispersed
define homogeneity
- little variabablity
define heterogeneity
- great variability
ex. wide range of age, attitudes
what is standard deviation r/t variability
- average deviation of scores in a distribution
- helps predict where people’s score may fall
define range r/t variability
- highest value minus lowest value
what is the P value
- the statistical calculation that shows a relationship between the variables that is unlikely to be due to change only
- describes the probability that there was some chance occurrence
what are some common bivariate statistical tests (3)
- t-tests
- correlation coefficients
- confidence intervals
what is a t-test
- tests the difference between two means
what are correlation coefficients
- describes intensity and direction on a relationship
- ranges from -1.0 to +1.0 where a score of 0 = no relationship