231 final exam- Quant designs and methods

chapter 9

define population:

the population is the entire group of interest of study.

how are the characteristics of a population specified?

they are specified through eligibility criteria, ex: consider the population of nursing students, do we include part time students in our population we want to study?

inclusion criteria: do they qualify as a member of the population, do they meet the characteristics

exclusion criteria: or do they not qualify as a member of the population, do they not meet the characteristics of the population.

define sample/sampling:

a sample is a subset of the population, we want to find a sample that is representative of the population

what is sampling bias?

is it when there is a systematic over or under representation of key characteristics of a population.

non probability sampling designs

the researchers select the participants in the study using non random methods.

it is less likely to produce representative samples of the population but it is the most common sampling approach in nursing.

probability sampling designs

it is a random selection from the population, meaning everyone in the population has an equal chance of being selected, (not the same thing as random assignment in RCT).

what are the 3 types of non probability sampling?

convenience, consecutive, purposive,

what is convenience sampling?

selecting the most convenient and available people, Ex: giving out questionnaires for people to fill out that are leaving the library.

it is the weakest form of sampling and has the highest risk of sampling bias.

what is consecutive sampling?

selecting all people from an accessible population normally over a specific time period or to a specific sample size, ex: studying ventilator associated pneumonia in an ICU unit and picking from eligible people who were admitted to an ICU in the last 6 months.

\

what is purposive sampling?

using knowledge we have about the population to hand pick sample members, this can lead to bias. researchers may decide to purposely to select people presumed to be knowledgable about the issues in the study, can be useful when the researchers want a sample of experts,

what are the 3 types of probability sampling?

simple random, stratified random, and systematic

what is simple random sampling?

a random selection of participants in a broad sampling frame, has a low risk of sampling bias. it is the most basic type of probability sampling,

ex: if nursing students at Stfx was the population, a student roster would be pulled and numbered, and then a online randomizer would be used to pick enough people to fill the needed sample size.

there is no guarantee of a representative sample, but randomly selecting people guarantees that differences between the sample and the population happen by chance.

what is stratified random sampling?

a sampling frame is divided into different strata and participants are randomly selected within each strata. the goal is promote a more representative sample.

what is systematic sampling?

involves selecting the every kth person from a list, this can make a similar result to simple random sampling.

ex: selecting every 10th person from the list to be a participant,

sample size

it is a major concern in quant research, generally the bigger the sample size the less risk of sampling bias, this alone can’t protect from research being poor, but a large non probability sample is better than a small one.

what is a power analysis?

it is a tool that researchers can use to estimate how large their sample size should be in order to adequately test a hypotheses.

how can we critique sampling plans?

the researcher should describe their sampling strategy including:

what approach they used

their target population

the eligibility they had to meet (inclusion exclusion criteria)

sample size (and the rationale on how they arrived to this being the needed number)

and a description of the characteristics of the sample

\
they should also ideally include info about response rates.

Questionnaires: what is it, what are the strengths and limits

questionnaires are when the participants complete the instrument themselves with either open or close ended questions.

a benefit being they less costly, advantage to be able to target people in a dispersed area, they also offer anonymity which can be helpful when collecting sensitive info.

a limit is they normally have low respond rates, not all people can respond to them ex: children, low response rates may also lead to bias.

\
ex: sending out a survey about how Stfx students find res life, online with close ended question options.

data collection in quant research:

\-existing records are important data source for researchers

\-self reports are the most common data collection approach

what are likert scales, benefits and limits.

a scale is allows people to rate things numerically,

a likert scale has several declarative statements and asks how much they agree or disagree with the statements.

a limit being they can lead to bias, people often pick choices that are social norms or have a tendency to express extreme values.

\
they can be used with most people which is a benefit,

Scale- visual analogue scales, benefits and limits

a scale allows people to rate something numerically, can be used with most people. but they can lead to bias as people often pick social norms or express extreme values.

the visual analogue scale consists of a line with two extreme limits, the person is then prompted to drag the arrow along the scale to what most aligns with their opinion.

observational methods: category systems, benefits and limits

observation requires the formulation of a system for accurately categorizing, recording, and encoding the observations.

\
a category system records events of interest that happen within a setting systematically,

ex: if we were studying children’s behaviour “strikes object” “throws toy”

category systems are the basis for constructing a checklist,

a limit is bias, the researchers own values and prejudices may lead to difference in which behaviours are viewed or not from different researchers.

observational methods: rating scales

rating scales, the researchers rate phenomena over a scale that they are observing,

also a limit being bias between different researchers.

bio physiologic measures- in VIVO

in vivo measurements are those preformed directly on living organisms like blood pressure, body temp.

bio physiologic measures: in VITRO

meaning chemical measurements like hormone levels or bacterial levels, they are analyzed by lab techs.

define reliability

the extent to which scores are free from measurement error and are consistent over time for people who have not changed

ex: is the thermometer we are using giving the same reading for the same person within a couple min or is it not reliable?

how can we test for reliability?

test retest: aka stability or reproducibility, extent to which the same scores can be obtained on repeated admin when trait being measured has not changed.

\
inter-rater: how reliably the measures reflect the attribute of the person being assessed, and is not related to characteristics of the viewer, important for observational measures.

\
internal consistency: When a measure has multiple components that produce similar measurement of the same trait.

\

how is reliability measured?

with correlation coefficients that range from 0.0-1.0 if its above 0.8 they are desirable.

define validity

the extent to which a measure is actually measuring the constructs it is meant to measure

ex: are we measuring a babies pain or are their facial expressions being caused by them having a bowel movement/gas.

how can we test for validity?

1. face, if the measure appears to be measuring the target, this is not the best way to test it.
2. content, extent to which the instruments content adequately captures the construct, often achieved through expert consultation and consideration of the literature on a construct.
3. criterion, extent to which scores on a measure correlate with a “gold standard” measure, this is not possible for all measures ex, if there is no gold standard available.
4. constructs, the relationship between the scores on a measure and those measuring other similar/related constructs ex: known group comparison.