Evidence Based Research Test One
Methods: Measures/ Analysis
Are the measures used consistent with the design? Are the analyses consistent with the methodology? Multiple regression - need multiple variables - how were these measured? Constant comparative analysis - did this get adequately described in the grounded theory report?
P-value
P-value: how likely we got that result by chance (possibility we will not get the same result, higher the better) - Percentage of time that results could have happened by chance - Use for interpreting the statistical significance - Usually reported p < .05 (read = less than a 5% chance) p < .01 (read = less than a 1% chance)
What does PICO stand for?
P: pop/patient I: Intervention Procedural touch Caring Assessment Catheterization C: comparison O: outcome
Probability Sampling Techniques: Stratified Random Sample
get everyone one in that element ex. Look at so many Asian students so many African American students ect.
Probability Sampling Techniques: Random Sample
go down list and go every third person
Art of Nursing
goes back to that aesthetic way of knowing, being able to utilize empirical knowledge (research findings) with personal and ethical and socio-political knowledge (intuition and experience).
Probability Sampling Techniques: Cluster sample
look at particular area (pick so many people from each category so most of pop is not from certain area- reduce bias)
Non-Probability Sampling Techniques: Matched Sample
look for particular variable
Probability Sampling Techniques: Systematic Sample
put it into computer and have it spit out every 3rd person
Non-Probability Sampling Techniques: Convenience Sample
take first group you get
Non-Probability Sampling Techniques: Snowball Sample
"I heard you are in this study, I wanna be in this study" (combination of convince)
Meta-Analysis
(combination of studies) (means researched a lot) - A quantitative research study of other studies - Must follow the steps of the research process, to include statistical analysis of the data
Systematic Review
(does nothing w/ data, overview of all articles about topic) - Review of the literature to synthesize the state of the science for an over-arching practice problem - Describe findings by clinically meaningful topics - what themes are identified? What is the general consensus?
t-distribution
- "t" value is a test statistic for differences in the means between two groups - Requires continuous variables Can be the differences between two different groups (i.e. men and women) Can be the differences in the same group at two different times (i.e. nursing students at the beginning of the nursing program and nursing students at the end of the program)
Theory
- A rigorous structuring of defined concepts and statements that present a systematic view of phenomena that can be used to describe, explain, predict, and control the phenomenon. - developed from purpose, concepts, relationships, structure, assumptions - Quan: based o theory (gives guidance) - Qual: developing theory
Anonymity v Confidentiality
- Anonymous: No one, including the researcher, can link study data to an individual - Confidentiality Ensures identity of individuals cannot be revealed to anyone else - Once a study is completed, must ensure the participants' anonymity - destroy any identifiers for potential for future work
Quantitative Research Designs: Non-Experimental Quasi-Experimental and Experimental
- Answer questions involving prediction and effects of manipulation - Difference: amount of control - Quasi: do not have total random control trail (have everything else), rare can one group get something and one cant get something - Experimental: have to have random control trail
Level of measurement: Interval (Continuous)
- Assign numbers in meaningful order with equal distances between points of measure - Zero point is not meaningful Temperature in Fahrenheit (0o does not mean "No temperature")
Levels of measurement: Ordinal (Categorical)
- Assign numbers so order IS meaningful - Intervals BETWEEN numbers are not consistent Education Level, Income categories
Levels of measurement: Nominal (Categorical)
- Assign numbers to categories - Ordering of numbers not meaningful 0 = male, 1 = female 0 = married, 1 = widowed, 2 = divorced, 3 = single - Give number so we can put in computer and calculate (always us to categorize info)
Variance
- Average of the squared deviations from the mean - average variation - The more scores differ from each other and the mean, the greater the VARIANCE
Consent v Assent
- Consent is given when a person is fully aware of the implication of participation - CONSENT is given by a guardian/parent because the participant is unable to provide consent - Assent is obtained from children (usually less than 18 years of age) unless emancipated or teen mother of a child
Problem statement
- Contains what will be examined EX. Physical, pain-related, cognitive, stress-related, affective, and spouse-related variables all contribute to differences in pain behaviors in subgroups of patients with fibromyalgia syndrome - Include the population of interest - Include the variables being studied Univariate- 1 Bivariate- 2 Multivariate- multiple - Identify independent vs. dependent variables I: intervention or constant - ex. Letting drs know when to take catheter out D: the outcome
Quantitative Research Designs: Non-Experimental Correlation
- Describe possible interrelationships among variables - is there a correlation in results? Between choice and their actual results? Results are better with picking than assigning
Descriptive Statistics
- Describes the characteristics of the data collected (need to know- race/ culture/ language) Number of participants- N= Male/female- who was in the study Ages- life span is v dif Working/Unemployed- pay? Access? Education Level- 8th grade level - Will be found in any research article
Background & Significance Section: Literature Review
- Focused summary of what has already been published on the topic - Provides a picture of what is already known, as well as what is not known - Sources should be current and relevant
Hypothesis
- Formal statement of the expected relationship between two or more variables - Translates the research questions into predictions of expected results or outcomes
Research Problem
- Includes the background and significance What is the current state of the science? What do we know? What do we need to know? Why is this important? - May include the theoretical framework - Should provide the over-arching research question or goal of the study
Quality Improvement / Quality Assurance
- Institution-based standards of care - Discuss what the standard is based on - collection of data from the practice-setting - summarize findings of practice-improvement-project
Background & Significance Section
- Introduction of the research - Should include: Sufficient explanation of the problem A literature review Theoretical or Conceptual Framework: gives ideas about concepts and how they connect (picture of concepts in the research/ what you are examining) Hypotheses / Research Questions: tells you what researchers believe they will find Purpose of the study (why do we need): guides/focuses question
Inferential Statistics
- Make inferences about populations from a smaller group of data (Make an inference, based on this info we suspect that this could happen) - Statistical tests allow you to make GENERALIZATIONS - Results intend to explain or predict variables Probability of outcome occurring T-test: 2 groups and have mean ANOVA: compare more than two groups Regression: look at multiple variables (look at high risk factors) Factor-Analysis: look at survey, do we have a ? about each topic interested in
Mean v Median v Mode
- Mean Average - Median Midpoint in a set of scores - Mode Value that occurs most frequently in a set of scores
Level of measurement: Ratio (continuous)
- Meaningful order/equal distances between points - DOES have an absolute zero point (Zero is absolute- continuous but has an end to it) Height, Weight, Number of years at University of Portland
IRB/Institutional Review Board
- Meant to protect the public - Members—FDA requires at least 5 Based on Type of Studies Being Review Lay Persons—at least one outside of the institution Member from each Profession Represented - Scheduled Meetings are Published - Initial Review—how study will be done and how participants will be protected - Report of Study Results/Progress Yearly - Report of Any Problems Immediately - Random Internal Audits to Insure Protocol - External Audit Can Occur
Evidence based practice
- Multiple resources and has been repeated (not just one article) and looks at how this can be applied to practice - Conscious and intentional use of research- and theory-based information to make decisions about patient care delivery
Quantitative Research Designs: Non-Experimental Descriptive
- Portray a phenomenon as accurately as possible via statistics instead of language - not changing anything or doing intervention, asking client to complete survey and interview client results: yes/no answers and getting mean value (quan) vs quotes (qual)
Chi-Square
- Presented in research as: χ2 - Non-parametric test of differences between 2 groups (do it or don't do it) - Instead of mean and SD, comparison of frequencies to support a hypothesis - Compare what expected v what actually say - Statistically significant: a difference does exist
Qualitative Versus Quantitative
- Qualitative What is the "Quality" of the experience? Theory-building or theory-generating Understanding an experience within a certain context Asking clients "tell me how that felt" "what was that treatment like" "tell me about your experience" - Quantitative How do we "Quantify" the experience? Conducted within established theory Understanding the nature of an experience in order to be able to control and predict Measure (put numbers to ideas)- how many times did someone feel out of control
Random Sample v Random Assignment
- Random Sample: A Probability Method All members of a population had the opportunity to be in the sample - Random Assignment: Method to reduce bias in sampling Can happen for either probability or non-probability Randomly assign participants to different groups
The Research Process (the steps)
- Step 1: Define the problem Why is it important? Who will this study help? - Step 2: Develop a plan (how) What is the method that answers the question? How will the study by conducted? - Step 3: Implement the study Able to follow the plan? If not - keep track of what changes occurred - Step 4: Analyze and interpret Conduct the statistical analysis Analysis could be part of the implementation Is the data saturated? If not - purposive sampling - Step 5: Dissemination of findings MOST important step - if the findings are not disseminated, they cannot be utilized! Even need to report if hypotheses not supported
Research Design
- The overall plan for acquiring new knowledge or confirming existing knowledge - Two purposes in selecting a design: 1) To plan an approach that will best answer the research question 2) To ensure rigor and validity of results
Methods
- This section gives you the "building blocks" of the study Who was studied? With what design?- quantitative or qualitative How was the data analyzed?
Theoretical Framework
- Underpinnings of the study - An organized and systematic articulation of a set of statements related to questions in a discipline Its purpose is to describe (set forth what is) explain (account for how it functions) predict (under what conditions it occurs) prescribe (under what conditions it should occur) (tells you what concepts examined, relationship between concepts) - Can be identified at the beginning of a Quantitative Study - May be identified at the end of a Qualitative Study - Are used in Qualitative Research to add Scientific Rigor - Can be developed through Qualitative Research
Quantitative Research Designs: Non-Experimental Case-Control
- Usually retrospective; compare one group with situation to another group without - one group does one thing and another group does another thing
Background & Significance Section: Theoretical or Conceptual Framework
- Will tell you much about the method of the study - Qualitative or Quantitative? - Does the design "fit" the research question or problem?
What hinders EBP?
- Workplace issues/attitudes Does the agency support EBP? Real world needs vs. this is how we were taught Agency philosophy regarding who is involved in EBP - Individual issues/attitudes Do you question or "go with the flow"? Do you know how to find and evaluate evidence when you need it? Can you/will you apply evidence to the clinical situation?
Approaches to finding answers
1. Consulting an authority 2. Using intuition and subjective judgments 3. Turning to experience 4. Reading nursing research
Evidence Based Practice Two Key Concepts of Nursing Research:
1. Must describe the systematic gathering of information 2. Must address a gap in the current knowledge Simply stating "this area is understudied" is not sufficient for determining a GAP in knowledge WHY is it important to know?
Sampling Ethics: 5 Rights
1. Right to self-determination 2. Right to privacy and dignity 3. Right to anonymity and confidentiality 4. Right to fair treatment 5. Right to protection from discomfort and harm
Results: Significance
A statistical term indicating a low likelihood that any differences or relationships found in a study happened by chance. Ensures the finding was not "by chance"
What promotes EBP? Use the Nursing Process
Assessment: Knowing when and how to question. Knowing how to collect the most relevant and/or best evidence Diagnostically: Knowing how to read and interpret the evidence Planning: Integrating all sources of evidence. Discerning if the evidence applies to this situation. Implementing: Making the practice change. Informing those in power of the need to make the change. Evaluating: Evaluate the practice decision or change - what worked? What didn't work? What is within your power to change?
Standard Deviation
Average deviation from the mean - Square root of the variance - The smaller the SD, the more representative the mean is of the sample - If you see a Mean reported, you SHOULD see a SD report, as well - Want results to be tight- do not want large spread
Purpose Study
Contains why study is conducted EX. The primary objective of the present study was to confirm the role of significant others and of physical and psychosocial variables in the presence of pain behaviors for FMS patients
Methods: Design
Describes the way in which the study was conducted Qualitative versus Quantitative Procedures Were the methods of gathering data consistent with the methodology?
Non-probability Sampling
Do not assure that everyone in population will have chance to be in study, more practical, plausible, or possible than probability
Understand Limitations
Do not rely only to understand entire study or make decisions about care Use as a guide: 1. Does this study address my clinical question? 2. Was the study situation similar to my clinical site?
Quantitative Research Designs: Non-Experimental Cohort Studies
Follow group of subjects longitudinally over time to describe or predict Ex. Drugs how they impact life and outcome, how stress leads to life outcomes and disease
Methods: Sample
How many subjects? Important to know for assessing the quality of the study design How were they recruited?- how participants were found (volunteer, incentive, required) From what setting?- reflect pop I'm providing care for? Does the sample represent the targeted population of interest?
Confidence Interval
If I do this over and over what is the possibility I will get results I am supposed to get - Provides the range of actual values in which 95 out of 100 values would fall - The smaller the CI range, without including zero, the more confident the test statistic reflects the "real world" ATI Comprehensive—CI of 95% passage of NCLEX—student score must be at least 74.7%
Hierarchy of Evidence
Level I: Evidence from a systematic review or meta-analysis of all relevant randomized controlled trials (RCTs), or evidence-based clinical practice guidelines based on systematic reviews of RCTs- most highly regarded of lit Level II: Evidence obtained from at least one well-designed RCT Level III: Evidence obtained from well-designed controlled trials without randomization Level IV: Evidence from well-designed case-control and cohort studies Level V: Evidence from systematic reviews of descriptive and qualitative studies (not as much control, cant control bias as much) Level VI: Evidence from a single descriptive or qualitative study Level VII: Evidence from the opinion of authorities and/or reports of expert committees
Non-Probability Sampling Techniques: Quota Sample
Meet quota
Informed Consent
Must ensure all 5 rights are observed Must provide full disclosure of risk vs. benefit Must inform of right to withdraw from study at ANY TIME Must NOT coerce participation in ANY way
Levels of theoretical thinking
No hypothesis in qual
Asking Questions
Past history of searching for literature Current strategies Asking the right question Background Question vs. Foreground Question
Conclusions / Discussion
Provides the researcher's interpretations of the results Usually provide implications for nursing practice SHOULD discuss the limitations of the study - what limits the findings? - guides future research and allows the reader to understand how the study may NOT be applicable to his/her population So what? Why implement or not implement?
Results
Provides the statistical findings of the study Describes the qualitative themes of the study Does not include any interpretation by the authors - just the "facts"
Phenomenology
Qualitative Research Design - lived experience (observing)
Grounded Theory
Qualitative Research Design - researcher trying to understand process Examine how A and C, how did they get to C (examine stages)
Historical
Qualitative Research Design Ex. Had to wear gloves due to HIV
One-tailed t-test:
Results are interesting only if they turn out in a particular direction (confident where it will go) - state hypothesis and state which direction it will go
Two-tailed t-test:
Results would be interesting in either direction (not sure which way it will go) - Running the statistical test with no "direction" of the hypothesis - Know there is a difference, but don't know how groups will differ i.e. Health of nursing students will differ between married versus single students: Don't know if single will be higher or lower than married
5 Steps of Evidence-Based Practice
Step 1: Ask the burning clinical question (why am I interested in this) Step 2: Collect the most relevant and best evidence Step 3: Critically appraise the evidence Step 4: Integrate all evidence with one's clinical expertise, patient preferences, and values in making a practice decision or change Step 5: Evaluate the practice decision or change
Probability Sampling
Strategies to ensure every member of population of interest has an equal opportunity to be in the study
Research
one single study/ systematic review (take all findings and out numbers together and run data on it)
Peer reviewed
reviewed by someone in your profession (looking at it scientifically)