Research Exam 3

Cluster sampling

A multi-stage frame is developed that includes a list of all areas that could be used in a study, such as cities, institutions, or organizations. These areas are known as clusters. A randomized sample is then drawn from the list.

Alpha

A test's probability of making a type I error. The point at which the results of statistical analysis are judged to indicate a statistically significant difference between the groups.

Analysis of variance or ANOVA comparison to other analysis methods

ANOVA is more flexible than other analyses in that it can examine data from two or more groups.

Alpha level for most nursing studies

0.05. This means there is a five percent chance that the researcher's conclusion could result in a type one error.

Scatterplots

A graph of plotted points on a horizontal or x-axis and a vertical or y-axis that shows the relationship between two sets of data. If no correlation exits, the plots are random. If a large correlation exists, the points concentrate near a straight line.

Systematic variation

consequence of selecting subjects whose measurement values differ in some specific way from the population. For example, if all subjects in a study examining healthcare knowledge have an IQ higher than 120, scores are likely to be higher than in a study with a wide variation of IQ's. This introduces bias from the start.

Unstructured interview

content controlled by study participant using open-ended questions.

Regression analysis example as a predictor

could be used to predict the length of stay in a neonatal unit based on the combined effect of multiple variables such as gestational age, birth weight, number of complications, and sucking strength.

Ordinal level of measurement

data are assigned into categories that can be ranked. Cannot be shown that the intervals between the categories are equal. For example, normal weight for height, overweight for height, or underweight for height.

t-Statistic

data collected at the interval or ratio level of measurement and a mathematical equation produces the t-statistic which is compared with t values in a statistical table that identifies critical values of t.

Administrative data

data collected for reasons other than research, such as demographics.

Secondary data

data collected from previous research and used by other researchers for their studies.

Recruitment of study participants

depends upon the design of the study. Important to recruit originally planned sample size because data analysis and interpretation of findings depend on adequate sample size.

Descriptive statistics

describes the sample or variables and includes mean, median, mode, range, standard deviation, and z-score.

Types of quantitative studies that require large samples

descriptive and correlational. Multiple variables may be examined and extraneous variables likely will affect responses.

Two categories of statistics in nursing research

descriptive statistics and inferential statistics

Conclusions

a synthesis of finding using logical reasoning. A conclusion is a formation of meaningful whole from pieces of information obtained through data analysis and finding from previous studies. It also considers alternative explanations of the data.

Example of type one error

a test that shows a patient to have a disease when in fact the patient does not have the disease.

Non-significant results could stem from what type of error

a type two error. The researcher or theory used by the researcher to develop the hypothesis is in error, resulting in a "false negative" or type two error.

Absence of zero point

absolute amount is not known. Zero is still a temperature. There cannot be an absence of temperature.

Indirect measures

abstract concepts, such as pain, depression, coping, self-care, and self-esteem.

A significant study is

accepted by others and referenced in literature.

The population from which the researcher selects the actual study sample is referred to as the

accessible population.

Factors associated with physiological measures, such as blood pressure, pulse, etc.

accuracy, precision, and error.

Reliability testing—homogeneity

addresses the correlation of each question to the other questions within the instrument. Do all of the items in the instrument measure the same concept or variable? Testing for internal consistency using Cronbach's alpha coefficient.

Test-retest reliability

administration of the same instrument to the same subjects under similar conditions two or more times. Assumption is that the attribute to be measured remains the same at the two testing times. Are the results consistent?

Significant and predicted results

agree with those predicted by the researcher and supports logical links developed by the researcher among the framework, questions, variables, and measurement tools.

Analysis of covariance or ANCOVA

allows the researcher to examine the effect of treatment apart from the effect of one or more potentially confounding variables. Common confounding variables include pretest scores, age, education, social class, and anxiety level.

Non-significant results

also called negative or inconclusive results, this result occurs when the study did not detect any differences or relationships between the groups. This doesn't mean that no relationship exists—it means that the study failed to find a relationship.

Sampling plan

also called sampling method, this defines the sample selection process.

Analysis of variance or ANOVA is reported as

an F statistic.

Target population

an entire set of individuals or elements who meet the sampling criteria. For example, all people with diabetes in the country.

Significant and unpredicted results

are opposite of those predicted and indicate flaws in the logic of both the researcher and the theory being tested. If valid, they are an important addition to the body of knowledge.

Highly controlled research setting

artificially constructed environment developed for sole purpose of conducting research. Used in experimental research. For example, a lab.

Positive correlation

as one variable increases in value, so does the other one (+1).

Negative or inverse correlation

as one variable increases in value, the other variable decreases (-1).

Homogenous sample

as similar as possible so as to control for extraneous variables. This decreases the ability to generalize findings.

Example of a positive correlation

as the number of hours clients engage in brisk walking increases, so does the muscle tone in the lower extremities.

Example of an inverse correlation

as the number of hours worked increases, the hours of leisure time decreases.

Number of variables and its effect on sample size

as the number of variables increases, the sample size may increase. The inclusion of multiple dependent variables creates a need for increased sample size in order to detect the differences between the groups.

Measurement variance effect on sample size

as variance increases, the sample size needs to increase.

Decision theory

assumes that all groups in a study used to test a hypothesis are components of the same population relative to the variables under study.

Purposive sampling

based on the researcher's judgement. Goal is to select appropriate participants to obtain in-depth information as the researcher finds the subjects and asks them to participate in the study.

Analysis of variance or ANOVA can look at what three variances

between-group variance, within-group variance, and total variance.

Chi-square test of independence is used

by nurse researchers interested in the number of participants or events that fall within specific categories. Only conveys the existence or non-existence of a relationship and not its strength.

Systematic variation in a study with a high acceptance rate or a low refusal rate

chance for systematic variation is less and the sample is more likely to be representative of the target population.

Exclusion criteria

characteristics that can cause a person or element to be excluded from the target population. For example, previous knee replacement surgery.

Reliability in measurement theory

concerned with how consistently the measurement technique measures the attribute of interest. Getting consistent results with the same instrument.

Reliability testing—stability

concerned with the consistency or stability of repeated measures of the same attribute with the use of the same scale or instrument. Same results obtained on repeated administration of the instrument.

Direct measures

concrete things such as oxygen saturation, temperature, and weight.

Likert scale

designed to determine the opinion or attitude of a study subject. Contains a number of declarative statements with a scale after each statement. Allows the individual to express how much they agree or disagree with a particular statement. Most common in nursing.

When the effect size is large

detecting it is easy and only a small sample is needed.

Measurement error

difference between true measure and what is actually measured by the instrument. There is no such thing as a perfect measure.

Two categories of directness of measurement

direct and indirect measures.

Five concepts of measurement theory

directness of measurement, measurement error, level of measurement, reliability, and validity.

Factors considered when analyzing significance of findings

does it make an important difference in people's lives, is it possible to generalize findings to a larger population, and do the findings have implications for disciplines other than nursing.

Factors that influence the adequacy of sample size in quantitative study

effect size, type of quantitative study conducted, number of variables, measurement sensitivity, and data analysis techniques.

Minimum acceptable power level is

eighty percent (80%).

Interval level of measurement

equal numerical distances between intervals. For example, temperature—the difference between 70 degrees and 80 degrees is the same as the difference between 30 degrees and 40 degrees. Absence of zero point.

Random variation

expected difference in values that occur when different subjects from the same sample are examined. As sample size increases, random variation decreases, improving representativeness.

Generalization is

extending findings from the sample under study to a larger population. This is why the sample must be representative of the population the study is generalized to.

Generalizing the findings

extends the implications of the findings from the sample studied to a larger population, or from the situation studied to a more general situation.

Statistical significance versus clinical significance

findings can have statistical significance but not clinical significance.

Sampling in qualitative studies

focuses on experiences, events, incidents, and settings rather than on people.

Sampling frame

for everyone in the accessible population to have an opportunity for selection in the sample, each person in the accessible population must be identified using sampling criteria. The subjects are then selected using a sampling method or plan.

Risks in developing conclusions

going beyond the data and forming conclusions not warranted by the data.

Ratio level of measurement

highest level of measurement. There is an absolute zero point, or an absence of property being measured. For example, weight, length.

Likelihood of generalization in inferential statistics is determined by

how big the sample is. The bigger the sample, the more confident you can be that the sample prevalence is close to the population prevalence.

Analysis of variance or ANOVA compares

how much members of a group differ or vary among one another with how much members of the group differ or vary from members of other groups.

Measurement sensitivity

how reliable was the tool? Was it a valid way to measure?

Reliability testing—validity

how well the instrument measures what it is supposed to measure. Complete validity is not possible. Researcher determines degree of validity.

Inferential statistics

identifies relationships or reaches conclusions about an entire population based on the results of a sample of that population.

Sample size and nature of topic in a qualitative study

if a topic is clear and easily understood, fewer subjects may be needed.

Sample size and quality of data in a qualitative study

if quality of the data is high, with rich content, fewer participants are needed to reach saturation.

t-Test significant difference is determined how

if the computed value is greater than or equal to the critical value, the groups are significantly different.

Significance of findings are associated with

importance to the nursing body of knowledge. May be associated with the amount of variance explained, control in the study design to eliminate unexplained variance and the ability to detect statistically significant differences or relationships.

Causes of type two errors

inappropriate methods, biased or small sample, internal validity problems, inadequate measurements, weak statistical measures, or faulty analysis.

A correlation near what number is indicative of a strong inverse correlation

negative one.

Four levels of measurement

nominal, ordinal, interval, and ratio.

When renewing a nursing license, nurses are asked to indicate their area of specialization. This datum is measured on which scale

nominal. It is just categories.

Observational measurements

observer watches the participant in a specific setting. Often used in qualitative studies. Limitation is that this is more subjective than other methods.

Range

obtained by subtracting the lowest score from the highest score. Very crude—is sensitive to outliers because it only uses to two more extreme scores.

A statistically significant finding means that

obtained results are not likely to be due to chance.

Suggesting further studies

occurs when the researcher gains knowledge and experience from conducting the study that can be used to design a better study next time.

Example of regression analysis

one might wish to predict the possibility of passing the NCLEX based on GPA in a nursing program.

Sample in a qualitative study

participants

Who should judge clinical significance

patients and their families, clinician, researcher, society at large. In any case, clinical significance is ultimately a value judgement.

Measurement and data collection strategies

physiological measurements, observational measurements, interviews, focus groups, questionnaires, and scales.

Two kinds of correlation are

positive and negative.

A correlation near what number is indicative of a strong positive correlation

positive one.

Power analysis

power is the probability that a statistical test will detect a significant difference that exists. This determines the risk of a type two error.

Clinical significance is related to

practical importance and relevance of the findings to clinical practice. There is no common agreement in nursing about how to judge clinical significance.

Questionnaires

printed self-report form designed to elicit information that can be obtained from subject's written responses. Can be distributed to large sample in person, by mail, or online.

Two types of sampling plan

probability and nonprobability.

Three things considered with reliability of measurement theory

stability, equivalence, and homogeneity.

Two types of interviews

structured and unstructured.

Sample size and scope of qualitative study

study with a broad scope requires larger sample than a study with a narrow scope.

Sample in a quantitative study

subjects

Two types of measurement error

systematic and random.

Specificity

the ability of an instrument to screen out those without a condition correctly. Highly specific tests have a low rate of false positives.

Sensitivity

the ability of the instrument to screen in or diagnose a condition correctly. High sensitivity has a low rate of false negatives.

Variables are measured with

the best possible measurement method available to produce trustworthy data that can be used in statistical analysis.

Inclusion criteria

the characteristics that the subject or element must possess to be part of the target population. For example, must be over age 18, able to speak and read English, and have a surgical replacement of one knee joint in the last year.

Limitations may limit

the credibility of findings and conclusions as well as generalizability.

Precision

the degree of consistency or reproducibility of the measurements.

Factor analysis aids in

the development of theoretical constructs and measurement scales. The researcher must explain why the analysis grouped the variables in a specific way.

Standard deviation

the square root of the variance; the average difference score. Numeric calculation that measures the average amount that each value in the data set differs or deviates from the mean.

Power analysis

the statistical method used to estimate sample size needs.

Mean

the sum of the scores divided by the number of scores being summed.

Normal curve

the theoretical frequency distribution of all possible values in a population. Bell-shaped curve. Mean, median, and mode are equal.

Systematic error

the variation in measurement values from the calculated average is in the same direction. For example, the scale is off by three pounds.

Random error

the variation in measurement values from the calculated average is without patter. For example, the date is entered incorrectly.

In analysis of variance or ANOVA, an F statistic equal to or greater to the appropriate table value means

there is a statistically significant difference between the groups.

Chi-square test of independence example

to determine if the infection rate in one nursing home is the same as the infection rate in another nursing home.

Decision theory rules were designed for what reason

to increase the probability that inferences are accurate.

Considerations for Data collectors

training process and interrater reliability.

Quality of diagnostic and screening tests yield four possible results

true positive, false positive, true negative, and false negative.

Bimodal

two modes exist.

Interrater reliability—testing for equivalence

two observers measure the same event at the same time. Two observers using the same instrument at the same time will yield similar results.

Alternate forms of reliability—testing for equivalence

two version of the same instrument are tested to compare for reliability. Two instruments, when completed by the same group of study participants, should yield equal results. For example, SAT.

Natural research setting

uncontrolled, real-life environment. Used in descriptive, correlational and qualitative research.

Visual analog scale

used to measure strength and intensity of an individual's subjective feelings or attitudes about symptoms or situations. A 100mm line with right-angle stops at each end is presented to the subject who records their response to a study variable by placing a mark through the line to indicate intensity of feeling. The researcher measures the line from the left end to the subject's mark.

Focus groups

used to study qualitative issues. Obtain participants' perceptions of narrow subject in a group interview sessions. Gives the group of a feeling of safety in numbers. Used for facilitating data collection with potentially threatening, uncomfortable research topics. Discussion helps to provide depth of data.

Summated scales

various items on most scales are summed to obtain a single score.

Example of factor analysis in a study of men's attitudes toward vasectomy

various themes of masculinity, loss of ability to reproduce, etc., should serve as the basis for constructing subscales because there are various themes to men's feelings about vasectomy.

Data analysis techniques

vary in capability to detect differences in data.

Interviews

verbal communication between the researcher and the subject, during which information is provided to the researcher.

Regression analysis is used

when one wishes to predict the value of one variable based on the value of one or more other variables. It describes how two or more variables are related.

Data saturation

when the researcher is detecting no new knowledge from additional subjects.

Sample attrition

withdrawal or loss of subjects from a study.

A correlation near what number indicates that there is no particular relationship between variables

zero.

Correlation range

negative one through zero and up to positive one.

Relationship between validity and reliability

Can have reliability without validity, but not validity without reliability. Able to consistently measure the same thing, but not measure what you think you are measuring. For example, the scale used to calculate weight has not been calibrated, but you consistently measure weight. The scale is not valid, but the consistency of the measurement is reliable.

Partially controlled research setting

Environment is manipulated or modified in some way by the researcher. Used in correlational, quasi-experimental, and experimental research.

Quality measurement methods

Improve the accuracy and validity of findings.

Systematic sampling

Members of the sample are selected at fixed intervals from a list, like every fifth person. It is important that the list not be compiled in an order that would create bias.

Nonprobability sampling plan

Not random. Not every element of the population has the opportunity to be selected. Decreases representativeness. Often used in nursing research because of the limited number of patients to draw from.

Probability sampling plan

Random. Each person or element has opportunity to be selected for the sample.

Network sampling

Subjects who have knowledge of a situation are identified by others in the same situation. Often uses social networking. Also called snowball sampling. Eligibility must be clearly defined and this process can be slow and costly. No guarantee of representativeness.

Purposes of statistical analysis

To examine the numerical data gathered in a study and to provide insight into the meaning of data.

Difference between a type one error and a type two error

Type one error occurs when a fire alarm going off indicating a fire when in fact there is no fire. A type two error occurs when a fire breaks out and the fire alarm does not ring.

Theoretical sampling

Used in grounded theory research to develop a selected theory. Sample is any individual or group that can provide relevant data. The sample is saturated when data collection is complete based upon the researcher's expectations.

Stratified random sampling

Used when the researcher knows some of the variables in the population are vital to achieving representativeness. Subjects are divided into subgroups, then randomly chosen proportionally from the different subgroups, which are often based on demographic variables such as age, income, gender, or race.

Consistency in data collection

Vital to accurate data collection. Important to maintain data collection pattern for each collection event as it was developed in the research plan. Deviations from the plan and its impact on the interpretation of findings should be noted in the report.

Example of a type two error

a blood test failing to detect the disease it was designed to detect, in a patient who really has the disease

Cronback's alpha coefficient

a calculation used to test for internal consistency and determine reliability and homogeneity.

Inference

a conclusion or judgement based on evidence. Judgements are based on statistical results.

Quota sampling

a convenience sampling technique with an added strategy to ensure inclusion of subjects that are of interest to the study. Helps increase representativeness and decrease bias.

Closely related variables are grouped into what

a factor.

Probability theory is expressed as

a lowercase p with values expressed as percentatges or decimal value, ranging from zero to one.

A high retention rate results in

a more representative sample of the target population and a more accurate reflection of reality.

Decision theory is usually expressed as

a null hypothesis. There is no difference between the groups. It is up to the researcher to provide evidence that there really is a difference.

Population definition

a particular group of individuals or elements who are the focus of research.

Physiological measurements

increased need for ways to measure the outcomes of nursing care has resulted in more nursing studies that include physiological measures such as electronic monitoring equipment, physical measurement methods such as blood pressure and pulse ox, and microbiological such as smears and cultures.

Power of the analysis technique

increases as precision in measurement increases. The power of the planned statistical analysis also dictates sample size—if it is weak a larger sample is needed.

Studies that obtain data from existing databases

increasingly used in nursing research, existing databases, sometimes from government data, contain large amounts of relevant information, larger samples, and result in lower costs. Quality of database used should be addressed in report, as well as type of database and validity and reliability of data. Data should address study purpose.

Variance

indicates the spread or dispersion of scores. The larger the variance, the larger the dispersion of the scores.

Elements

individual units of the population and sample. For example, a person or event. When the element is a person, they are referred to as participants or subjects.

Questionnaire versus interview

information obtained is similar, but questions have less depth. Data collector cannot probe for information and subject can't elaborate or ask for clarification. Less opportunity for bias due to questions being presented in a consistent manner.

Reliability testing—equivalence

involves the comparison of either two versions of the same instrument or two observers measuring the same event at the same time.

Convenience sampling

involves using the most conveniently available people as study participants. This is the weakest type of sampling and it is difficult to control for bias. Available subjects may be atypical of the general population, but the benefit is the study will receive the largest number of subjects in the shortest amount of time.

Effect size

is the extent to which the null hypothesis is false. Just how false is it?

Effect

is the presence of the phenomenon being studied.

P value

is the probability that an outcome would occur by chance. In other words, this is the percentage of chance that errors exist. The smaller the number, the more significant the finding because the less likelihood there is for error.

Analysis of variance or ANOVA F-distribution table

is used to determine the significance of the calculated F statistic

An acceptable p value in nursing

less than five percent.

Four components of power analysis

level of significance, sample size, power and effect size.

Sampling criteria

list of characteristics essential for eligibility or membership in the target population. For example, sampling criteria for a study on the effects of early ambulation on length of hospital stay for adults having knee replacement surgery may include must be over age 18, able to speak and read English, and have a surgical replacement of one knee joint in the last year, no previous joint replacement surgery, and no dementia or debilitating chronic muscle diseases.

Error

may be caused by environment, user, subject, equipment, interpretation.

Focus group considerations

may sort participants into smaller groups with common characteristics. Need to select an effective moderator to keep discussion on track. The setting should be relaxed and comfortable. High-quality recordings should be made.

Directness of measurement

measurement begins by clarifying what is to be measured.

Z-score

measures how far from mean an individual value is, expressed as the number of standard deviations by which it differs from that mean.

Measures of central tendency

mode, median, and mean.

Inferential statistics is for research in which

more is needed than to describe a sample. This method infers or predicts how likely it is that the data can be generalized to a wider population.

Scales

more precise means of measuring phenomena than questionnaires.

Simple random sampling

most basic. Achieved by randomly selecting elements from the sampling frame. For example, drawing names from a hat, having a computer randomly select subjects, or using a table of random numbers.

Statistical inferences

must be made with great care.

Three types of research settings

natural, partially controlled, or highly controlled.

Data collection

process of acquiring subjects and collecting data for study. Steps are specific to each study and depend upon design and measurement techniques. Includes training data collectors, recruiting study participants, implementing intervention, collecting data in a consistent manner, and protecting validity of the study.

Standardized scores

provide a way to compare scores in a similar process. Common standard is a z-score.

Three types of sampling in qualitative studies

purposive or purposeful, network, and theoretical.

Types of quantitative studies that require smaller samples

quasi-experimental and experimental. As control in the study increases, sample size can decrease. Instruments are more refined, therefore stronger validity and reliability.

Pearson product-moment correlation is represented by the letter

r

Measures of dispersion

range, variance, standard deviation, standardized scores, and scatterplots. These examine how scores vary and are dispersed around the mean.

Three types of scales

ratings scales, likert scale, and visual analog scales.

Use of statistics in nursing practice

reading and critiquing published research, examining outcomes of nursing practice by analyzing data, developing administrative reports, analyzing research done on clinical site, and demonstrating a problem or need and conducting a study.

Unexpected results

relationships found between variables that were not hypothesized and not predicted from the framework being used. These results can be useful in theory development, modification of existing theory, and development of later studies. However, they must be evaluated carefully as the study was not designed to examine these results.

The researcher often makes suggestions for further studies that logically emerge from the current study, such as

replications or repeating the design with a different or larger sample, different measurement methods, testing a new intervention, or forming a hypothesis and strategies to further test the framework used.

Heterogeneous sample

represents a broad range of values and diversity. This increase the ability to generalize the findings of the study.

Structured interview

research controls the content by asking specific questions. Responses are entered on an instrument.

Structured observations

researcher defines what is to be observed and how the observations are to be made, recorded, and coded. Category system is used to organize and sort behaviors and events observed. Checklists are used to indicate whether a behavior occurred. Rating scales are used to rate the behavior or event.

Sample attrition as noted in the study

researcher should provide the subject's reasons for withdrawing.

Significance of findings - researcher expectations

researchers are expected to clarify in the report the significance of the study findings.

Control in the study design

researchers build control into the study plan to minimize influence of extraneous variables. Final report should include controls implemented in the study, as well as issues that arose during the study and how they were managed.

Limitations to statistical analysis

restrictions or problems in a study that may decrease the generalizability of the findings or a combination of theoretical and methodological weaknesses.

Sample size in qualitative studies

sample size is adequate when data saturation occurs.

Representativeness of samples in quantitative and outcomes studies

sample, accessible, and target populations are as alike in as many ways as possible in order to enable generalizations.

Subjects who participate in a study of patients with inflammatory bowel disease are described as the

sample.

Ratings scales

scale that lists an ordered series of categories of a variable. Assumed to be based on an underlying continuum. Numerical value assigned to each category. Examples are faces pain rating scale or numerical zero through ten pain scale.

Existing healthcare data consists of

secondary data and administrative data.

What is sampling

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

How many factors may be identified within a data set

several may be identified.

Five types of results in statistical analysis

significant and predicted results, non-significant results, significant and unpredicted results, mixed results, and unexpected results.

t-Test tests for

significant differences between two samples.

Types of probability sampling

simple random sampling, stratified random sampling, cluster sampling, and systematic sampling.

Sample size and qualitative study design

some studies require more interviews, family participation, etc.

Unstructured observations

spontaneously observing and recording, these may lead to structured observations.

Analysis of covariance or ANCOVA versus analysis of variance or ANOVA

the effects on study variables are statistically removed by performing regression analysis before performing ANOVA. This removes the effect of differences among groups that is caused by the confounding variables (age, education, social class, etc). This allows the effect of the treatment to be examined more precisely.

Probability theory is used to explain

the extent of a relationship, probability of an event occurring, and the probability that an event can be accurately predicted.

Accuracy

the extent to which the instrument measures what it is supposed to measure.

A type two error (or error of the second kind) is

the failure to reject a false null hypothesis. A false negative.

A type one error (or error of the first kind) is

the incorrect rejection of a true null hypothesis. A false positive. Usually a type I error leads one to conclude that a supposed effect or relationship exists when in fact it doesn't.

Example of no correlation between variables

the last two digits of the social security number are not in any way associated with body temperature.

In decision theory, to test the assumption that there is no difference between groups, a cutoff point is selected before analysis. This is referred to as

the level of statistical significance, or alpha.

Likelihood ratios

the likelihood that a positive rest result is a true positive and that a negative test result is a true negative.

Nominal level of measurement

the lowest level. All data will fit into categories that are not ranked in any way. For example, data is gender, categories are male or female.

Implications for nursing

the meanings of conclusions for the body of nursing knowledge, theory, and practice, this section is based on the conclusions but provides more specific suggestions for implementing the findings in nursing.

Median

the midpoint or the score at the exact center of the ungrouped distribution.

Mixed results

the most common result. One variable may uphold predicted characteristics, whereas another does not, or two dependent measures of the same variable may show opposite results. This result may be caused by methodology problems, or may indicate a need to modify the existing theory.

Sample retention

the number of subjects who remain in and complete the study.

Mode

the numerical value or score that occurs with greatest frequency.

Acceptance rate

the percentage of subjects who consent to be in the study.

Refusal rate

the percentage of subjects who decline to participate in the study.

Accessible population

the portion of the target population to which the researcher has reasonable access to select sample. For example, all people with diabetes within a particular health plan.

Pearson product-moment correlation tests for

the presence of a relationship or correlation between two variables, called the bivariate correlation.

Measurement

the process of assigning numbers or values to individual's health status, objects, events, or situations using a set of rules.

The ability to generalize a study is influenced by

the quality of the study and the consistency of the study's findings.

Factor analysis examines

the relationships among large numbers of variables. It disentangles those relationships to identify clusters of variables most closely linked and sorts variables according to how closely related they are to other variables.

Descriptive statistics is for research in which

the researcher does nothing to a single group of subjects except measure them. Data is collected and no inferences are made.

Researcher's responsibility regarding limitations

the researcher should identify limitations and discuss how they may have affected the findings and the conclusion.

Findings are

the results of the study that have been translated and interpreted—a consequence of evaluating evidence.

Levels of measurement

the rules for assigning numbers to objects so that a hierarchy in measurement is established. When critically appraising a study, the level of measurement achieved should be noted. Researchers try to achieve the highest the level of measurement. Ratio is highest and nominal is lowest.

When the effect size is small

the sample size needs to be increased in order to detect the difference between the two groups.

What is a sample

the selected group of people from which data is collected for a study. Should be representative of the population identified in the study.

Sampling in research may be defined as

the selection of a subset of a population to represent the whole population.

Whether a correlation is strong enough to be statistically significant depends upon

the size of the actual sample and the level of significance chosen by the researcher.