Chapter 10 - Sampling & Data Collection in Quantitative Studies Flashcards
What is the population?
The entire group of interest
The collection of elements/cases in which a researcher is interested in
What is a sample?
a section of a portion of the population to represent the entire population
A subset of population elements→ elements are usually humans
What is eligibility criteria? Inclusion vs. exclusion criteria
specific characteristics to define the population
Inclusion criteria: characteristics that make one a member of the population
Exclusion criteria: characteristics that cause one to be excluded from population
Target population
the entire population in which a researcher is interested (ex. all diabetics in the US)
Accessible population
portion of the target population that is accessible to the researcher (ex. all diabetics at one hospital)
Strata
mutually exclusive segments of a population based on specific characteristics
(Ex. RNs in the UK can be separated by GENDER, Years of experience)
–> Strata are often used in sample selection to enhance the samples representativeness
Representative Sample
one whose characteristics closely approximate those of the population
Sampling Bias
systematic overrepresentation or underrepresentation of some segment of the population in terms of key characteristics when the sample is not representative
–>can be risky to apply it to the population because it could be very incorrect
A representative sample is most easily achieved with… (3 things)
- Probability sampling – sampling at random
- Homogenous populations – more alike than different
- Larger samples – people in sample are more likely to have the characteristics you are looking for
Nonprobability Sampling
researchers select elements by nonrandom methods in which every element usually does not have a chance to be included
- rarely produce representative samples
- MOST research samples in nursing are nonprobability
Convenience Sampling
selecting the most conveniently available people as participants
-problem: atypical population, bias –> weakest form of sampling
Quota Sampling
identify population strata and figure out how many people are needed from each stratum –> ensures diverse segments are accurately represented
-participants are a “convenience sample” from each stratum
Consecutive Sampling
recruiting ALL people from an accessible population over a specific time interval or for a specified sample size
- longer sample time = less bias due to seasonal fluctuations
- best choice when their is “rolling enrollment” in population
–>ex. ICU patients admitted over 6 months, or the first 250 patients
Purposive Sampling
based on the belief that researchers’ knowledge about he population can be used to hand-pick sample members
- can lead to bias (based on purposeful judgement of the researcher)
- useful when you want a sample of experts
- mostly qualitative
Probability Sampling
involves random selection of elements form a population - all elements have equal chance of being selected
- random sampling
- only viable method of obtaining representative samples, can easily estimate sampling errors (ex. average age of population) and sample values (ex. average age of the sample)
- often impractical
Random Sampling
- used on probability sampling
- ->each element in the population has an equal, independent chance of being selected
Simple Random Sampling
most basic probability sampling
- researchers establish a sampling frame (list of population elements –> ex. student roster for sampling students at MU, pulling names out of a jar)
- no guarantee of a representative sample, but there is no researcher bias (sample of CHANCE)
Stratified Random Sampling
population is first divided into two or more strata, from which elements are randomly selected
- ->ex. sex, ethnicity
- aim to enhance representativeness
- only works if stratifying variable is available
Systematic Sampling
selection of every “kth” case from a list (ex. every 10th person on a patient list)
-divide population into sampling interval and then select sample from different intervals
Sample Size
the number of study participants in the final sample
LARGER is better = less sampling error
-Sample size adequacy needed for quantitative research
Power analysis
helpful in estimating how large their samples should be for testing their research hypothesis
–>when predicted differences are small - a large sample is needed
What are the two issues to assess when critiquing a sampling plan?
- Has the researcher adequately described the sampling strategy?
- Has the researcher made good sampling decisions?
Ideal description of the sampling strategy
- The type of sampling approach used
- The population under study and eligibility criteria for sample selection
- The sample size, with a rationale
- A description of the samples main characteristic (age, gender, clinical status)
- The number and characteristics of potential subjects who declined to participate
How do you know if the researcher has made good sampling decisions?
- Response rates: the number of people who participated in a study relative to the number of people sampled
- Nonresponse bias: differences between participants and those who declined to participate (response bias)
- ->In longitudinal studies attrition (dropping out) bias should be reported
Data Collection: New data vs. old data
New data – data collected specifically for research purposes (primary data)
Existing data – records, historical data, existing data sets (secondary analysis)
Self-Report
participants’ responses to questions posed by the researcher
- ->best when researchers know in advance exactly what they need to know and can frame appropriate questions needed to obtain needing information
- most common in nursing studies
- direct observation may be used
Closed-ended questions
response alternative are prespecified (ex. yes or no)
- good closed ended questions are difficult to construct but easier to analyze, more efficient
- may be superficial
Advantages of Questionnaires
- less costly
- can be dispersed anywhere easily (especially over internet)
- possibility of anonymity
Advantages of Interviews
- more than questionnaires
- response rates are higher (don’t refuse to talk to an interviewer as much)
- interviews are feasible with most people (many cannot fill out a questionnaire - children, blind, elderly)
- interviewers can produce additional info through observation of responders behavior/living situation
What is a scale?
a device that assign a numeric score to people along a continuum (ex. scale for measuring weight)
–>social-psychological scales quantitatively discriminate among people with different attitudes, perceptions, and psychological traits
Likert Scale
several declarative statements (items) that express a viewpoint on a topic
–>express how much you agree or disagree with a statement
Summated Rating/Composite Scale: sum of item scores
Visual Analog Scale (VAS)
used to measure subjective experiences (ex. pain, dyspnea)
- a straight line with end anchors that are labeled as the most extreme limits (no pain vs. worst pain ever)
- should be 100 mm in length = score of 0-100
Social Desirability Response Set Bias
a tendency to misrepresent attitudes or traits by giving answers that are consistent with prevailing social views
Extreme Response Set Bias
a tendency to consistently express extreme attitudes or feelings (ex. strongly agree) leading to distortions because extreme responses may be unrelated to trait being measured
Acquiescence Response Set Bias
a tendency to agree with statements regardless of their content by some people (yea-sayers) or disagree with statements independent of the question’s content (nay-sayers)
What areas do observational researchers have flexibility?
- Focus of observation - broad (events) or narrow (specific behaviors - facial expressions)
- Concealment - may not tell subjects they are observing (reactivity: behavior distortions because they know they are being watched)
- Method of recording observations - paper and pencil recording or sophisticated methods (ex. video, audio, computers)
Category System
represents a method of recording in a systematic fashion the behaviors and events of interest that transpire within setting
Non-exhaustive: all observed behaviors are classified into one category
Exhaustive: many behaviors in many categories
Rating scale
observers rating phenomena along a descriptive continuum (occurrence + qualitative aspect of it)
Time sampling vs. event sampling
Time Sampling: selection of time periods during which observations will occur (ex. every 20 seconds at 2 min intervals)
Event Sampling: selection of integral behaviors/events to observe, more for infrequent events
Biophysiologic Measures
–> used for wide variety of purposes: basic biophysiologic processes, exploring ways in which nursing actions/interventions affect physiologic outcomes, product assessments, etc
In Vivo: performed directly within or on living organisms (ex. BP, temp)
In Vitro: gathered b extracting biophysiologic material from them and analyzing it (ex. hormone levels, bacterial counts, tissue biopsies)
Advantages of Biophysiologic Measures
- relatively accurate/precise
- objective (two nurses would read same BP on monitor/two monitors would read the same)
- pts cannot easily distort measurements
- provide valid measures of targeted measures (thermometers measure temperature, not blood volume)
Semantic Differential Scale
• Continuum between 2 adjectives
• Continuum varies from 5 to 9
Ex. My nurse is..
MEAN ——————-NICE