Statistics for Nursing and allied health

Evidence-based research
p2

Cochrane Collaboration
p2

description
exploration
explanation
prediction and control
p3

Descriptive
Inferential
p3

Numerical or graphical summaries of data, including charts, graphs and simple summary statistics such as means and standard deviations, used to describe characteristics of a problem sample.
p3

statistical techniques (chi-square test, the t-test, the one-way ANOVA) that allows conclusions to be drawn about the relationships found among different variable in a population sample.
p3

Studies whose primary purpose is descriptive and explorative simply describe situations and events.

Ex: What is the marital status of people in the US?

p3-4

Primary purpose is explanatory elucidate the relationship among variables...more complex than descriptive.

Their questions and lines of inquiry are often based on established theories from the research literature.

Explanatory studies depend on inferential questions: Are women who are sedentary during third trimester of pregnancy more likely to have a cesarean section than women who exercise regularly during the third trimester?

studies do not necessarily attempt to establish causality but often attempt to understand how variable are r/t each other.
p4

Seek to determine which variables are predictive and to determine causality (one event causes another to happen)

Prediction and control studies are typically quasi-experimental or experimental studies in which researchers introduce an intervention.
p5

random selection

an intervention

one or more control groups that do not receive an intervention

random assignment of study participants to either the control/intervention group
p5

Are similar to true experimental designs except that they lack one or more of the following:
random selection into the study
a true control group
random assignment to the intervention or control group

This design is often used in the health-related research

Quasi-experimental typically use inferential statistics
p5

1. Undertaking a research project w/out reviewing the existing literature on the subject
2.Collecting data w/out a well-defined plan, hoping to make sense of it afterwards.
3.Trying to fit meaningful research questions to existing data
4.Defining terms in general or ambiguous language
5.Failing to base research on a sound theoretical foundation
6.Failing to make explicit and clear the underlying assumptions
7.Failing to recognize the limitations of the approach
8.Failing to anticipate rival hypotheses that would account for findings and that challenge interpretations and conclusions.
p6

Mistakes are often made when there is no study plan or when the plan is insufficiently detailed.
p6

It is a written presentation of how the researcher is going to obtain and analyze the numerical data needed to answer the research questions.

keeps the analysis focused and relevant
it serves as the basis for the introduction and methods section of research papers after the data have been collected and anaylzed.
study plan can also serve as the basis for the first 3 chps of a dissertation or thesis.
p6

1.Statement of the problem and its significance
2.Theoretical or conceptual framework
3.Research questions to be answered by the study
4.List of hypotheses to be tested
5.Definitions of key terms and variables
6.Description of the research design
7.Description of the sample and how it was obtained
8.Description of the planned statistical analysis
9.Statement of assumptions and limitations.
10. Dissemination plan
p7

The purpose of the study, what the study is trying to solve.

A short description of the significance of the problem, no more than 2-3 sentences

THE STATEMENT OF PURPOSE IS THE GUIDING FORCE BEHIND THE ENTIRE RESEARCH PROJECT.
p7

A study plan needs a theoretical or conceptual framework on which research questions and hypotheses are based.
This framework presents a structured way of thinking abt the interrelationships of the variables, how they relate to one another.
p7

Empirical research is a way of gaining knowledge by means of direct and indirect observation or experience. Empirical evidence (the record of one's direct observations or experiences) can be analyzed quantitatively or qualitatively. Through quantifying the evidence or making sense of it in qualitative form, a researcher can answer empirical questions, which should be clearly defined and answerable with the evidence collected (usually called data). Research design varies by field and by the question being investigated. Many researchers combine qualitative and quantitative forms of analysis to better answer questions which cannot be studied in laboratory settings, particularly in the social sciences and in education.

1. Andersons model of Health Care Use
2. Theory of Unpleasant Symptoms
3. Theory of Planned Behavior
p8-9

It postulates that the use of health services is a function of the perceived need for care

enabling factor/ need factor ex. model to study condom use among adolescents having easy access to condoms is viewed as enabling.
p8

This theory postulates that symptoms do not occur in isolation, meaning that the effects of symptoms do not occur in isolation, meaning that the effects of symptoms that occur at the same time are multiplicative and not simply additive.
p9

A theory often used in the research of individual health behavior and behavioral intentions.

The performance of any behavior depends on behavioral beliefs, normative beliefs and control beliefs.
p9

It is important to ground research questions in the existing literature.

Questions should be clear abt the relationships that are expected.
p9

What is the level of intention to engage in physical activity among a group of adults that recently joined a fitness facility?
p10

Does attitude toward exercise affect participation in physical activity?
p10

Hypothesis is a tentative prediction or explanation of the relationship b/w 2or more variables.

The purpose of a hypothesis is to translate problem statements and research questions into predictions of expected outcomes.
p10

In a research plan, it is important to clearly define key terms and variables. Terms are best defined when they first appear so that readers do not initially make assumptions about definitions only to discover later that different definitions apply. It is especially important to define terms that readers outside the field of study may not understand.
p11

Are those that are manipulated (ie by the intervention) or that may affect the outcome.
Ex: age, income, preexisting health conditions.
p11

Are those that are expected to change in response to the intervention such as health status, # visits to the doctor, costs of inpatient stay
p11

Every study plan must include a description of how the data will be collected.

When collecting original data, the plan must describe in detail exactly how this will be accomplished.

Research design is the art and science of conducting studies that maximize reliability and validity of the findings and minimize bias.
p12

observational studies
quasi-experimental
experimental studies
p12

Are those in which a phenomenon is simply observed and no intervention is instituted.
They are appropriate when the purpose of the study is descriptive or correlational.
p12

Cross sectional
Longitudinal study
p12

Study that involves the collection of data at one point in time, the sample can be random or use intact groups or convenience samples.
p12

Are designed to collect data at more than one point in time. they are appropriate when casual links need to be established b/w independent and dependent variables.
p12

The researcher is an active agent in the quasi-experimental and experimental study designs.
p12

Same:
*Involve measurements taken during at least 2 separate pts in time, pretest and posttest.
*involve an intervention
*comparision group (control)

Differ:
Quasi is not random and exper. is.

**Evidence from experimental studies is considered to be stronger than evidence from quasi-experimental studies.
p13

T

p13

The process of selecting a portion of the population to represent the entire population.

Can be random or nonrandom
p13

In general, studies attempt to find participants who represent all members of the population of interest cause it is generally impossible to gather the data from the entire pop. So its done thru sampling.

Important to state exactly what the target population is.
p13

Cleaning the data
describing the sample
inferential statistics used to test each hypothesis.
p14-15

Involves making certain that all the variables have valid and usable values and to check the variables with missing data...examining for valid values, unusual values, large amts of missing data and adequate variability.

For example: female=1, male=2 but there is a 3 or if the pts weight is 890lbs
p14

Describe the sample characteristics using descriptive statistics. it is common practive to create tables.

p15

List the inferential statistics that will be used to test the hypotheses.
p15

Every study has these, all of which must be stated explicitly.
Assumptions: statemts that are taken to be true even if the direct evidence of truth is either absent or not well documented. Ex:it is assumed that participants are honest.
Limitations: weaknesses that potentially limit the validity of the results. Ex. using only a limited non-random sample, sample sizes fewer that 100
Dislimitations: boundries in which the study was deliberately confined, EX. study focuses on adults and not adolescents/adults.
p15-16

A study is not truly complete until the knowledge gleaned from the work is disseminated for use. The most important part of a research project is sharing the knowledge obtained from the project.
p16

Variables measure the different characteristics or parameters of a given sample. A variable is any characteristic that can and does assume different values from person to person. for examp: demographic variables describe basic charc of human pop such as gender, age, ethnicity, marital status.
To be considered a variable it is critical that every1 in a given sample doesn't have the same value. For exam: gender can assume 2 values (male/female), in a study of pregnant women, gender doens't vary and can't be considered a variable.
p20

Categorization
rank ordering
interval ordering
numerical scoring
p23

nominal (categorization)
ordinal (rank ordering)
interval (interval ordering)
ratio (numerical scoring)
p23

Is the lowest form of measurement, the numbers are simply used as labels, representing categories or characteristics and there is no order to the categories.

Ex.: gender, religion, ethnicity, etc.

Each value can be assigned a #, Males=1, females=2...numerical assignmt is arbitrary cause assigning #2 to females and 1 to males changes nothing.

2 category variables are called dichotomous.
p23

Can be placed in a meaningful numerical order (lowest to highest), but there is no info abt the specific size of the interval b/t the different values and there is no true zero. Numbers are simply used to put observations in rank order.
Ex: it is possible to rank military personnel from lowest to highest, private, corporal, sergeant, lieutenant but it is impossible to say anything meaningful abt how much greater a corporal is than a private.

**All subjective rating scales are considered ordinal, including satisfaction scales, etc.
p23

Can be placed in a meaningful numerical order and have meaningful intervals b/w values cause the units of measure are in equal intervals.
*no true zero
*EX: Fahrenheit temp
*This scale isn't used by healthcare workers expect IQ scales and some psychological measurmts.
p24

*Can be placed in meaningful numerical order, have meaningful intervals and have a true zero.
*Ex: wt, ht, bp, pulse rate, age, income and # of children.
p24

central tendency
dispersion
shape
outliners
p24

The values that best represent the middle of a distribution.

They provide info abt the values that are most typical.
p24

Describes the extent to which the values of the variable are spread out around the measure of central tendency.

Standard deviation, interquartile range and range are all measures of dispersion
P24

Describes how the values of the variable are distributed (symmetricall or asymmetrically) around the measure of central tendency.
This is typically determined visually using graphical methods such as histograms/stem-and-leaf plots.
p25

Values that do not fit the pattern of the rest of the values of the variable, they are either much larger/smaller than the rest, they stand out cause they are unusual.
p25

descriptive statistics
frequency distributions
graphical displays
p25

It is used to describe and summarize data to make them more meaningful. These include measures of central tendency, dispersion, shape and skewness.
*It is often presented visually using frequency distributions and graphical displays such as histograms and bar charts.
*Overall this allows readers to better understand the pop on which the studies were focused.
p25-6

It provides a way of organizing data using a table format.
*They allow readers to grasp the basic charac of the distribution of the variables quickly.
*When the data are organized into a table format, it is much easier for readers to glean info abt central tendency, dispersion and outliners of the variable of interest.
*Is useful when describing ordinal, interval and ratio data.
p26

Most studies are conducted using only a sample of data from the population of interest. The assumption made is that the results obtained from the sample are a valid respresentation of the whole population. However, an estamate of a mean made from sample data does not provide the exact value of the true mean of the population. One way to assess how good an estimate is obtained from the sample is to compute confidence intervals around the estimate of the mean.
CI are typically constructed as either 95% or 99% CI. A 95% CI is interpreted as follows: if the study were repeated over and over again, drawing different samples of the same size from the same population, 95% of the time the estimated mean would fall within the CI. Statistics book p71

The program of adolescent youth that was discussed in the previous section had 62 participants. However, there are many more adolescents in the city of Buffalo that could have participated; these 62 participants represent the larger population of all adolescent girls who could of enrolled in the program. Then it would be reasonable to ask the following question: Was the population used in the study representative of a good population?
Statistics book p71

The actual probability of getting the obtained results or results even more extreme. The smaller the p-value, the more statistically significant (i.e., the less likely the result is due to chance)
p412

The specific level of a p-value that is defined as "statistically significant" is called the alpha-level ( -level). This value is defined by the researcher before any statistical tests are conducted. Common -levels used are .10, .05 and .01. An -level of .10 means that for a result to be significant, it cannot occur more than 10% of the time by chance. Similarly, an -level of .05 means that the result cannot occur more than 5% of the time by chance, and an -level of .01 mans that it cannot occur more than 1% of the time by chance. p85