Research Applications Flashcards
What are the 8 steps in research?
- identify a relevant and important topic
- develop well considered research question
- research question leads to a hypothesis
- prepare research protocol
- organize materials and methods
- collect and analyze data
- study results and make decisions
- study designs and checklists
How do you identify a relevant and important topic?
review published research literature related to the topic
What is a well-considered research question?
who, what, how; clear, simple statement in a few words, in a complete grammatical statement
What is a hypothesis?
a prediction of a relationship
How is a hypothesis often expressed?
as more than or less than; not equal to
what is a null hypothesis?
no relationship in population of data; any difference is result of sampling error; often has equal to expressed
What does a research objective do?
defines the study’s purpose
A good hypothesis should be ___.
feasible, interesting, novel or innovative, ethical, and relevant
What is the PICO format?
population, intervention/exposure, comparison, and outcome
What is the research protocol?
methodology to solve the problem
What is PRISMA?
PRISMA randomized controlled trials: an evidence based minimum set of items for reporting in systematic reviews and meta analyses
What does PRISMA focus on?
reporting of reviews evaluating randomized trials, but can also be used as a bases for reporting systematic reviews of other types of research, particularly evaluations of interventions
What is MOOSE?
systematic review-observational MOOSE; meta-analysis of observational studies in epidemiology (MOOSE) group
List the parts of the research report.
- abstract
- general introduction
- review of existing literature
- methodology
- results
- discussion
- conclusions
- implications
What is an abstract?
condensation of final report; has purpose of study, questions asked, scope and method, summary of conclusions
What is included in the general introduction?
objectives, definitions, background, limitations, order of presentation
What is the review of existing literature?
a summary of different points of view
What is included in the methodology?
statement of hypothesis, discussion of methods used
**What is included in the results section of a research report?
specific lab, clinical, objective or subjective findings
What is the discussion section of a research report?
interpretation of results, comparison with other studies; may be combines with results
Which section of a research report may the discussion section be combined with?
the results section
What does discussing the results address?
the research question, objective, and hypothesis; places results in context with existing science
What are conclusions in a research report?
brief summary of results; may have recommendations
What are implications in a research report?
how the information might be applied in practice
What is descriptive research?
describes state of nature at a point in time; provides baseline data and monitors changes over time
Descriptive research generates hypotheses regarding what?
determinants of a condition or disease
Does descriptive research prove cause and effect?
no; it establishes associations among factors, but does not allow causal relationships to be determined
List the types of descriptive research.
qualitative, case report, surveys, correlation/ecological studies
What is the purpose of qualitative research?
to explore a phenomenon of interest as a prelude to theory development; often preceded other research
When does qualitative research take place?
it often precedes other research
How is data collected for qualitative research?
through interviews, observations, questionnaires; may use focus group (Delphi)
What is the Delphi method in research?
use of a focus group
What is a case report/study/series?
report of observations on one subject or more than one subject
What does a case report help to identify?
variables important to the etiology, care, or outcomes of a particular condition
What does a case report describe?
it describes quantitatively the experiences of a group of cases with a disease or a condition in common
What are surveys and how are they used?
research designed to describe and quantify characteristics of a defined population; defined time frame; pinpoints problems
What are correlation or ecological studies?
studies that compare frequency of events (or disease rates) in different populations with the per capita consumption of certain dietary factors (ex. correlation between fish consumption and breast cancer incidence)
What is analytical research?
tests hypothesis concerning the effects of specific factors of interest and allows causal associations to be determined (can prove cause and effect); includes clinical trials, follow-up studies, and case control studies
Can analytical research prove cause and effect?
yes
What are the types of analytical research?
experimental model, quasi experimental design, cohort studies, case control studies, and cross sectional studies
What is the experimental model?
uses experimental and control groups; target populations are randomly chosen to be in either group
What is randomization?
dividing people into treatment or control groups without bias
**Does the experimental or control group receive the treatment in the experimental model?
experimental group; control group does not receive the treatment but may receive a placebo
What is a placebo?
gives the aura but not the substance of a service, removing the possibility of the Hawthorne effect
What is the Hawthorne effect?
a positive response due to attention that participants receive
When is a program considered successful in an experimental model?
differences are computed between the two groups; successful if the experimental group has improved more than the control group
Why is the experimental model difficult to run?
- not enough people for the control group
- may not feel it is ethical to deny a service
**What is the quasi-experimental design?
time series- series of measurements at periodic intervals before the programs begins and after it ends; shows whether the measurements before and after a program ate a continuation of a previous pattern or whether they indicate a noteworthy change
What does the quasi-experimental design show?
whether the measurements before and after a program ate a continuation of a previous pattern or whether they indicate a noteworthy change
What is a cohort?
any group whose members have something in common
What are cohort studies?
group of people who have something in common are followed over time; ex. cohort of healthy people followed through time to see if they develop a specific disease
Cohort studies are also sometimes called ___?
incidence studies; they track the frequency of new cases (newly diagnosed) of a disease
??Define incidence?
??rate of newly diagnosed
??Define prevalence?
??total number of
Cohort studies are carried out over what period of time?
long period of time (longitudinal), and prospective (future oriented)
How are retrospective cohort studies conducted and what is looked for?
use existing data; look back for relationship between exposure factors and outcomes
What are case control studies?
focus on specific disease; those with the disease are compared with a group without the disease, but otherwise similar in characteristics; both groups recall past behaviors, to study how the groups differ
**What are cross-sectional studies/prevalence?
one time data collection counting all of the cases of a specific disease among a group of people at a particular time; **a snap-shot look at one point in time
What does a cross-sectional study/prevalence look at?
one point in time; describes current, not past or future events
What does IRB stand for?
institutional review board
The IRB is under ___.
FDA
What is the IRB?
a committee established to review and approve research involving human subjects, to ensure it is conducted within all ethical and federal guidelines
What are other names for the IRB?
(IEC) Independent Ethics Committee; (ERB) Ethical Review Board; (REB) Research Ethics Board
What is relevance or validity?
ability to measure phenomenon it intends to measure
**What is internal validity?
tests whether the difference between the two groups is real (has the experimental group really performed differently)
What is external validity?
tests whether or not a generalization can be made from the study to a larger population
**What is analysis of variance?
tool used to evaluate validity
**What does ANOVA stand for?
analysis of variance
**What does analysis of variance ask?
whether the difference between samples is a reliable one that would be repeated; are there one or more significant differences ANYWHERE among the samples?
**When is ANOVA used?
when several products compete against one another
**What does ANOVA compare?
the variance between groups with the variance within groups
What is reliability?
consistency or reproducibility of test results; test then retest later-are results similar?
What are parallel forms and how is reliability determined?
two separate but similar forms of the same test at the same time; reliability is determined by the degree to which the sets of scores coincide
What are split halves and how is reliability determined?
divide the test in half; reliability is determined by the degree of similarity of results
What is precision?
amount of variation that occurs randomly
Less random variation results in ___ precision in the measurement and ___ reliability.
greater; greater
When is specificity and sensitivity used?
If protocol involves screening for a particular condition
What do specificity and sensitivity evaluate?
the cut-off value being used
What is sensitivity?
the proportion of afflicted individuals who test positive
What is specificity?
proportion of non-afflicted identified as non-afflicted
What are variables?
characteristics that may have different values from observation to observation
What are nominal variables?
non-ordinal; variables that fit into a category with no special order
What are examples of nominal variables?
gender, race, marital status, present or absent
What are rank order variables?
ordinal; observations compared with each other and put in order, perhaps from best to worst, or state of disease from 1 to 4
What are numerical discrete variables?
data with numbers (numbers of clinical visits)
What are numerical continuous variables?
Underlying continuous scales (blood pressure)
Dependent variables are ___. Independent variables are ___.
outcomes; what you manipulate
___ variables are outcomes; ___ variables are what you manipulate in your study.
dependent; independent
Treatments for diseases are ___ variables.
independent you can change the treatment to affect the disease
Name the independent and dependent variables in the following sentence. Effect cholesterol levels have on heart attacks.
Independent- cholesterol levels
Dependent- heart attacks
?What is probability?
?each segment of the population will be represented in the sample
What is randomization?
select a sample from the whole population so the characteristics of each of the units approximates the characteristics of the whole population
What is non-probability?
no way of forecasting that each element in the population will be represented in the sample
What is convenience or accidental?
take units as they arrive on the scene- no attempt to control bias
What is quota?
select units in the same ratio as they are found in the general population
What are measures of central of tendency?
the center of any mass of data
**What are the three measures of central tendency?
mean, median, and mode
What is the arithmetic mean? How is it calculated?
simple average; total the values of all observations and divide by the number of observations
What is the median? How is it calculated?
Midpoint; arrange observations from low to high and count the number of values; median is the value at the midpoint; if there is an odd number of numbers the median will be the exact center. If there is an even number of numbers, the median is the average of the two numbers closest to the center
What is the mode?
most frequently occurring value; prediction most likely to be right
What are measures of dispersion?
how values are distributed about the mean
What are types of measures of dispersion?
range and standard deviation
What is the range?
the difference between the lowest and highest values in the distribution
How is the range calculated?
subtracted lower from higher value; based only on extreme values
What is standard deviation?
indicates degree of dispersion about the mean value of a distribution
How is the standard deviation calculated?
square root of the sum of the squared deviations of each value from the mean, divided by the number of observations
Describe the curve of a normal distribution?
as it falls away from the peak on either side the slope is convex (bulging outward) at point of inflection it become concave (bulging inward) as slope begins to level off
Distance between the ___ and the ___ on either side is equal to the standard deviation.
mean; point of inflection
Distance between the mean and the point of inflection on either side is equal to the ___.
standard deviation
**About ___% of all observations in a normal distribution lie within 1 SD of the mean.
68% (2/3)
**About 2/3 (68%) of all observations in a normal distribution lie within ___ SD of the mean.
1 SD of the mean
What is the range within 1 SD of the mean called?
M + 1 SD of the mean (mean + 1 SD)
___% of the observations lie outside the range?
32% (16% lie below (-1 SD) and 16% lie above (+1 SD))
___% lie within about 2 SD either side of the mean.
95%
Where does the mean lie?
at the top of the curve
What are correlations?
relationships between varying types of data
The closer point are to the line, the ___ the degree of the linear relationship.
stronger
The ___ point are to the line, the stronger the degree of the linear relationship.
closer
What is the linear correlation coefficient?
r; measures the degree to which the points in a scatter diagram cluster about a straight line
The value of r is always between ___ and ___.
-1 and 1
r = 1 when ___.
all points lie exactly on a straight line with a positive slope
r= ___ when all points lie exactly on a straight line with a positive slope.
1
r = -1 when ___.
all points lie exactly on a straight line with a negative slope
r= ___ when all points lie exactly on a straight line with a negative slope.
-1
**The closer r is to 1 or -1, the ___ the points tend to cluster about the line, and the ___ the degree of the linear relationship.
closer; stronger
**The closer r is to ___ or ___, the closer the points tend to cluster about the line, and the stronger the degree of the linear relationship.
1 or -1
The closer r is to 0, the ___ the points will be from the line.
more dispersed
The closer r is to ___, the more dispersed the points will be from the line.
0
if r = ___, there is no linear relationship
0
if r = 0, is there a linear relationship?
no
Describe the strength of correlation if r =0.0-0.2
very weak
Describe the strength of correlation if r =0.2-0.4
weak, low
Describe the strength of correlation if r =0.4-0.7
moderate
Describe the strength of correlation if r =0.7-0.9
strong, high
Describe the strength of correlation if r =0.9-1.0
very strong, very high
Describe the strength of correlation if
a. r =0.0-0.2
b. r =0.2-0.4
c. r =0.4-0.7
d. r =0.7-0.9
e. r =0.9-1.0
a. 0.0-0.2: very weak
b. 0.2-0.4: weak, low
c. 0.4-0.7: moderate
d. 0.7-0.9: strong, high
e. 0.9-1.0: very strong, very high
With perfect positive correlation r = ___.
+1.0
With perfect negative correlation r = ___.
-1.0
Describe the slope when r = 1
positive slope, upward to right
Describe the slope when r = -1
negative slope, upward to left
What is clinical significance?
a change of difference in outcomes that somebody cares about; the outcome must be relevant for patient care, public health, or the field of study. the change must be statistically significant, not due to chance.
The p value represents what?
level of significance
**The lower the p value, the ___ the significance of your results
higher
**The ___ the p value, the higher the significance of your results.
lower
**What is the level of significance if p ≤ .05?
significant differences, results are reliable
**What is the level of significance if p ≤ .01?
very significant difference, more reliable results
**What is the level of significance if p ≤ .0.0001?
very, very significant, reliable results
**What is the level of significance if p > .05?
not very significant difference; not reliable results
- *What is the level of significance for the following p values:
a. p ≤ .05
b. p ≤ .01
c. p ≤ .0.0001
d. p > .05
a. p ≤ .05: significant differences, results are reliable
b. p ≤ .01: very significant difference, more reliable results
c. p ≤ .0.0001: very, very significant, reliable results
d. p > .05: not very significant difference; not reliable results
What does the p value show?
how strong or weak the evidence is in support of a hypothesis
Most will not accept results as statistically significant unless p___.
p<0.05
The ___ the p value, the higher is your confidence that the effect you observed is real.
smaller
The smaller the p value, the ___ is your confidence that the effect you observed is real.
higher
**What is a double blind study?
removes bias from research; neither the researcher nor the subject knows which group is receiving the treatment and which the placebo
**Define mortality.
rate of death
**Define morbidity.
state of disease
Define variable of interest.
what researchers are observing
Define population of interest.
describes a group about which the observations are made
What are descriptive statistics?
they summarize and describe aspects of a set of data
**What are inferential statistics?
techniques that allow conclusions to extend beyond an immediate data set
What are questions that inferential statistics answer?
- What is the probability that the results can be applied to a larger group?
- What can you infer from the results of your study?
What is a non-parametric test?
a test that does not depend on a normal distribution
What are dichotomous scores?
only two events are possible (ex. heads, tails)
What are continuous scores?
scores measured on a continuous scale
**What is a pilot study?
a scaled down version of the larger investigation; practice implementation; includes every step in the study, but is done on a small test group.
What can a pilot study help to determine?
whether a clinical trial, as planned, is feasible; are goals realistic and attainable?; will the study plan work?
What is a focus group?
a method of attaining information about a target group; small group who talk about the beliefs, opinions, problems; contributes attitudinal data
What is chi square?
χ2; tests whether or not there is a real difference between categories
When is the chi square test used?
with attributes that have more than 2 catgories
What does the chi square test compare?
the frequency with which we’d expect certain observations to occur with the frequency that actually occurred
What is a t test?
it tests the significance between the means of different populations; tests the null against alternative hypothesis
In a t-test, if the probability value is equal to less than the level set for significance, the ___ is rejected in favor of the ___.
null hypothesis; alternative hypothesis
In a t-test, if the probability value is ___, the null hypothesis is rejected in favor of the alternative hypothesis.
equal to less than the level set for significance
What is a histogram?
a block diagram whose blocks are proportional in area to the frequency in each call or group (frequency distribution of data)
What does a histogram summarize?
data from a process that has been collected over time
What is the EAL?
evidence analysis library; evidence based guidelines are developed by conducting a systematic review and then using the conclusion of the review to develop practice-based guidelines
Who summarizes the best available evidence for the EAL?
subject matter experts and trained analysts
The AND Evidence Analysis Library assists in what?
answering questions that may arise during the provision of nutrition care.
What is a line graph?
shows frequency on vertical scale and method of classification on the horizontal scale
What are bar charts?
show measurement only on the vertical axis; bars are arranged horizontally or vertically in ascending or descending order
What is a pie chart?
chart in which pieces add up to 100%
What is meta analysis?
a formal, defined system combining results of numerous small studies to increase the strength of belief in the observed effect
What are requirements for a study to be included in a meta analysis?
similar design, defined inclusion and exclusion criteria, published peer-reviewed studies
What does RCT stand for?
randomized clinical trial
What is the gold standard for clinical nutrition research?
randomized clinical trial with comparison placebo control group
Randomized clinical trials are best for evaluating what?
medical treatments; intervention with one or more treatments
What is parallel design RCT?
participants are randomly assigned to a particular treatment group and remain on that treatment throughout the study
What is a crossover design study?
each participant serves as his own control
What is a two period crossover design?
each participant would receive either intervention or control (A or B) in the first period, and the alternate treatment (A or B) in the second period
What is the major advantage of a crossover design?
variability is reduced because the measured effect of the intervention is the difference in that participant’s response to the intervention and control
The decrease in variance of a crossover study design allows use of a ___ sample size.
smaller
Are crossover study designs generally longer or shorter?
longer; but each is “exposed” to all treatments