quiz 4 research ch 11,12, 14

threats to internal validity

Chance History Maturation Instrumentation Regression toward the mean Attrition (Mortality) Sensitization (Testing or Practice Effect) Selection Interaction of > 2 Threats

statistical conclusione validity

Extent to which true relationships between variables are detected by statistical analyses How accurate are results?

The study population is _____________________________ A.The number of elements from which information is obtained. B.Findings based on the information obtained. C.The group of interest for the study from which the sample is selected from. D.The way elements are selected for the sample.

c

random variation

Existing natural differences, normal variation Expected differences between individuals in same sample

convenience sampling

"Accidental" Eligible, accessible subjects are recruited until desired number reached Allows for multiple obvious, subtle, and/or unconscious biases Easy, fast, inexpensive Researcher must acknowledge biases and discuss ways in which the results may have been affected

confidentiality

Researchers must make every effort to prevent anyone outside of the project from connecting individual subjects with their responses

beneficence defintion

do no harm Includes the right to freedom from harm and discomfort Includes the right to protection from exploitation Maximize benefits & minimize potential harm

internal validity

Extent to which results are true WITHIN the study itself

full board human research

Greater than minimal risk or is not eligible for expedited review (does not fit into one of the 7 categories for expedited research) Must be reviewed by a convened meeting of the IRB

3 major ethical principles involved in protecting research

beneficence justice respect for person

study population

is a group from which sample is taken

four types of review

non human research exempt from review expedited IRB review full IRB review

type 1 error

rejecting the null hypothesis when it is actually true

methods of random assignment to groups

sealed envelope coin flip draw

IRB duties

1. Approve, disapprove, or modify research activities covered by regulations 2. Conduct continuing review of research involving human subjects 3. Observe the consent process & research & verify changes 4. Suspend or terminate approval of research

Justice definition

Fair & equitable treatment Selection of participants fair & non-preferential Includes the right to privacy Vulnerable subjects are protected

population

The entire set of elements that the researcher is interested in studying; All who meet sampling criteria

Random sampling is also called_______________________. A.Availability sampling B.Probability sampling C.Prospect sampling D.Probation sampling

b

'Test/retest' and 'parallel forms of the same test' are_________________________________. A.Internal consistency procedures to verify the validity of a measure B.Internal consistency procedures to verify the reliability of a measure C.External consistency procedures to verify the validity of a measure D.External consistency procedures to verify the reliability of a measure

d

Informed consent includes_______________________________. A.Explaining the purpose of the research B.Explaining how participation will affect the participant C.Ensure participants understand the type of information being collected D.All of the above.

d

sample size

refers to number of elements in sample

sample statistics

results found in sample

face and content validity

Each question or item has a logical link with an objective Each aspect should have adequate representation Content is the extent to which the statement goes

external validity

Extent to which results are true outside of the study in target population Also called Generalizability How far can we generalize the results of the study? Can we apply the results to those outside of the study? Small sample size or non-representative sample limits generalizability Cannot generalize to populations other than those from which sample was drawn

methods of random sampling

Fishbowl: Put all names in bowl & pick until you get desired number List of Random Numbers Random Number Generator (computer)

construct validity

How truly study measures reflect theoretical construct of interest based on statisticalprocedures

voluntary informed consent

Individuals have right to determine own actions & freedom to make own decisions

purposive sampling

Judgmental or theoretical Actively and consciously seeks out certain people to be in the sample Used often in qualitative research and in instrument development typical, extreme cases experts

systematic sampling

May be random or non-random Uses an ordered list of subjects/elements A random starting point should be selected, & then every kth element is included in the sample k = (# on list divided by # needed in sample) May over or under represent some group if do not use list of all possible subjects Each selection is not independent of the other Problem if some periodic variation exists in the list

too small sample size

May lead to greater sampling error Subjects may be exceptions, not representatives May yield non-significant findings

too large sample size

May yield statistically significant result with no clinical meaning Large sample without attention to representativeness may result in biased sample & false results

power analysis

Method to reduce the risk of Type II error Type II error: Accepting the null hypothesis when it is actually false This means you really had significant results but your study /statistics did not have enough POWER to detect them Common problem in nursing research is insufficient power

Expedited Human Research (non-exempt)

Must be minimal risk Research must fit into one of 7 categories (i.e., blood draws, surveys, chart reviews, etc) Reviewed by a single member of the IRB

cluster sampling

No one master list exists but "clusters" of the population are known to exist Researcher: Identifies the clusters Selects random sample of clusters Selects random sample from within each cluster Problem may arise if there is a lack of precise standardization of criteria for the variable under study across the clusters

systematic bias

Over or under representation of a subgroup or attribute of the population- "sampling bias" Consequence of selecting subjects that differ from the target population in significant ways

respect for persons

People have right to make their own decisions without pressure, coercion, or control from others Involves understanding and ability to evaluate risks and benefits based on full information Includes right to full disclosure of what will be done & why: Adequate info & No deception

predictive and concurrent validity

Predictive is how far the instrument can predict the outcome Concurrent is how well an instrument compares with a second assessment concurrently done

simple random sample

Requires a master list of the entire population of interest May use a number table, computer generated number list A complex strategy - not simple or haphazard No way to predict who will be selected; each choice is independent of any other choice; each member of the population has an equal chance of being selected

network smapling

Snowballing Researcher finds a few subjects & asks them to recommend others for study Asks for referrals from initial subjects for others with similar experiences/characteristics Useful in studies of vulnerable populations or with sensitive issues

quota sampling

Used to recruit members of known subgroups within the population Identify sub-group and number of subjects needed for each group Recruit subjects to meet the quota (not done randomly, but usually by convenience)

A code of ethics has been developed for conducting research in general. A.True B.False

a

Face validity refers to____________________________________. A.A logic link between research instrument and objective B.Statistical procedures establish the contribution of each important factor C.How well an instrument compares with a second assessment concurrently done D.The degree to which an instrument can forecast an outcome

a

In qualitative research the saturation point determines the sample size. A.True B.False

a

In quantitative research the sample should be unbiased and representative of the population. A.True B.False

a

Researchers must ensure that the extent of harm or discomfort in research is not greater that encountered in everyday life. A.True B.False

a

Systematic sampling ______________________________ A.Needs a list of all elements in the study population. B.Uses the researcher's judgment as to the ability of the respondents to contribute to the research. C.Is based on convenience in accessing the sample population. D.Is none of the above.

a

Validity is the ability of a research instrument to measures what it is designed to measure. A.True B.False

a

Validity provides answer to the research question for which it was undertaken by using appropriate methods and procedures. A.True B.False

a

Which of the following is a procedure for verifying the internal consistency of the reliability of a measure? A.The split-half technique B.Parallel forms of the same test C.Test/retest D.None of the above

a

n quantitative research 'internal validity' is an alternative for judging which of the following in qualitative research? A.Credibility B.Transferability C.Dependability D.Confirmability

a

type 2 eror

accepting the null hypothesis when it is actually false

population parameters

are specific numbers describing characteristics of the population

A research tool is 100% accurate if affecting factors such as wording of the questions, moods and interactions between interviewer and interviewee are controlled? A.True B.False

b

A researcher wishes to conduct a survey of research staff at the University of Kent, to investigate their views about proposed changes to the library opening times. She obtains a list of all of the university's research staff, and selects 100 people from the list using a random number table. What method of sampling method is this? A.Systematic sampling B.Simple random sampling C.Quota sampling D.Cluster sampling

b

Construct validity refers to_____________________________. A.A logic link between research instrument and objective B.Statistical procedures establish the contribution of each important factor C.How well an instrument compares with a second assessment concurrently done D.The degree to which an instrument can forecast an outcome

b

In qualitative research 'transferability' is an alternative for judging which of the following in quantitative research? A.Internal validity B.External validity C.Reliability D.Objectivity

b

Which is an ethical issue to consider relating to the researcher? A.Maintaining confidentiality B.Inappropriate use of information C.Seeking informed consent D.Providing incentives

b

Which of the following is least likely to be a stakeholder in research? A.The research participants or subject B.The publisher of a research journal C.The researcher D.The funding body

b

Collecting information with the expressed willingness of the participation is also called_______________. A.Providing incentives B.The possibility of causing harm to participants C.Seeking informed consent D.Maintaining confidentiality

c

It is ethical to__________________________ A.Give a present before data is being collected. B.Collect sensitive information even if it may upset participants. C.Ensure confidentiality of the information collected. D.Collect information without the knowledge of participants.

c

Quota sampling______________________________________ A.Needs a list of all elements in the study population. B.Uses the researcher's judgment as to the ability of the respondents to contribute to the research. C.Is based on convenience in accessing the sample population and is the non probability version of stratified sampling. D.Starts with a few individuals who then identify others to be included in the sample.

c

Which of the following is not a non-random sampling method? A.Quota sampling B.Accidental sampling C.Cluster sampling D.Convenience sampling

c

In a survey of hospital out-patients' views about the quality of service they receive from a hospital, researchers wanted to canvas an equal amount of women and men. During one particular week 200 women and 200 men were asked to fill in a questionnaire. What method of sampling is this? A.Expert sampling B.Cluster sampling C.Systematic sampling D.Quota sampling

d

major issues in sampling

deciding on sampling plan (who, what, when, how, and how many) ethical principles and constraints in human subject research

definition of medical ethics

is a system of moral principles that apply values and judgements to the practice of medicine

sampling error is affected by two factors

random variation systematic bias

accessible population

those who meet criteria and are available to researchers

Using information improperly is considered ethical in most professions. A.True B.False

b

to be considered informed consent three elements are necessary and one can not be missing

sufficient information competence voluntariness

target population

to whom findings will be generalized

Which of the following is a factor affecting reliability of a research instrument? A.The wording of the questions B.The physical setting C.The respondent's mood D.All of the above

d

non probability sampling

used in most healthcare studies less likely to be non biased types: convenience, quote, purposive, network

professional responsibilities

Make sure all patients approached to participate in a study are able to give informed consent Make sure all patients participating in a study have given informed consent and are aware of all study procedures Support patients' decision not to participate, to participate, or to withdraw from a study Contact attending physician & investigator if there are any patient reactions that could be a result of study participation Contact appropriate personnel & IRB if there are any concerns regarding the ethical aspects or human rights of study participation

stratified random sample

Researcher knows that subgroups exist within the population that will affect the sample's representativeness Strata = Mutually exclusive subgroups within the population Define strata, then randomly sample from within each stratum Associated with the lowest sampling error May be proportionate or disproportionate (weighting may be used to adjust)

members of IRB

Scientists, Different professions; Non-scientists; Community representatives; Representatives of specials populations; Lawyers; Clergy

sampling defintion

Selecting a group of people, events, behaviors, or other elements with which to conduct a study

sampling frame

ia a list of each element in study population

anonymity

project cannot link individual responses with participants responses

purpose of IRB

protect human subjects