Exam prep week 11 (Mixed methods research clinical governance) - wk 12 (Descriptive data analysis and findings) Flashcards
What is mixed methods research?
“Research in which the investigator collects and analyses data, integrates the findings and draws inferences using both qualitative and quantitative approaches or methods in a single study”
(Tashakkori & Creswell 2007:265)
Studies that have used two designs, qualitative and quantitative
Offers a different approach
Not limited to constraints of one or the other
Journal Of Mixed Methods Research
Methodological Triangulation (Pluralism)
Triangulation : usually applied to qualitative research
Reduces error/increases rigor
Different methods of data collection used in same study
For example: interviews + participant observation + documentation + focus groups etc
Now used to denote single study using combination of research designs/paradigms
Triangulation (more commonly used term)= pluralism = Mixed Methods
Becoming much more common/popular especially in nursing research
Not always labelled as such
Terminology/Rationales Associated with Mixed Methods
Triangulation Completeness Off-setting weakness & providing stronger inferences Answering different research questions Wider explanation of findings Broader illustration of data Hypothesis development & testing Instrument development & testing
REMEMBER!!
The appropriate research design is the one that will best answer the research question
Value of Mixed Methods
Potential for more complete & comprehensive research opportunity
Can give additional perspectives & insights beyond scope of single design
Weaknesses of one method may be counter-balanced by strengths of another
Limitations of Mixed Methods
Complex
Time consuming
Involved
Resource-intensive
Knowledge required of researchers- both qualitative & quantitative knowledge
Understanding & acceptance by research community needed
Action Research
Narrows gap between research and implementation of results
Research in the real world situation
Occurs in a spiral/cycle design
May use any type of research methodology
Emphasis on continual improvement
Limitation: may be weak experimental design
Delphi Technique
Uses expert opinion on a clinical practice problem
Non-empirical approach (ie no data collection)
Useful when experimental approach not feasible
Limitation: only represents opinion
Clinical Guidelines
Provide recommendations on clinical management
Generate national/international consensus on management principles
Allow for application of Research and EBP relating to a specific area of clinical practice
Are not mandatory
Tools to guide clinical decision making
Tools to guide clinical decision making eg standards, policies, & procedures, algorithms, clinical pathways, clinical guidelines
Assist to make appropriate decisions about patient care to result in best patient outcomes
Historically developed by individual/groups of experts
No process to determine validity/reliability
Evidence based practice has led to clinical guidelines being systematically developed
Based on best available research evidence
Definitions
Algorithms: clinical guidelines on flowchart
Clinical pathways: document essential steps in a clinical process eg COPD pathway
Clinical guidelines: systematically developed statements to assist clinical and patient decisions
Policies: written plans of an organisation’s official position eg Medication Administration Policy
Definitions (cont.)
Procedures: series of formal steps for performing specific tasks
Protocols: rigid, prescribed statements
Standards: accepted discipline-based principles for patient care processes
Policy
“A document that describes the organisation’s purpose or standard for a given customer process or issue, the expected outcome, guiding principles, roles and responsibilities, definition of terms within the document and references. Compliance with policies is mandatory”
(“Definitions of policy related documents within WA Health,” n.d.)
Procedure
A document that generally supports a policy by describing an instruction that clearly prescribes the actions of each step of a process to be taken and by whom.
(“Definitions of policy related documents within WA Health,” n.d.)
Characteristics of Effective Clinical Guidelines
No internationally recognised framework but general world-wide agreement
Key qualities for guidelines to be effective
Should be considered when appraising existing national guidelines before adapting to local situation and when developing new guidelines
NH&MRC Nine Guiding Principles for Effective Guideline Development
Validity Reproducibility Representativeness Flexibility Adaptability Cost-effectiveness Applicability Reliability Usefulness
Clinical Governance
“A system through which organisations are accountable for continuously improving the quality of their services and safeguarding high standards of care by creating an environment in which excellence in clinical care will flourish”
(Scally & Donaldson 1998:61 in Courtney & McCutcheon 2010:116)
What Do Clinical Audits Measure?
Monitor use of interventions/care received by patients against agreed standards
Assess effectiveness: does an intervention do what it is intended to do?
Need evidence to determine if intended outcomes were achieved
Audits show success/failure of intended outcomes
Identify departure from “best practice”
Descriptive vs Inferential Statistics
Descriptive statistics:
Allow researchers to describe, organise and summarise raw data
Inferential statistics:
Allow researchers to estimate how reliably they can make predictions and generalise their findings based on the data
Results Section of Research Papers
Summarise findings with two major goals:
- To describe/explain phenomenon of interest
- To predict aspects related to that phenomenon
Data Analysis in Qualitative Research
Research question always kept in mind
Data collected are formally interpreted
Data analysis is:
deliberate
considered
systematic
Qualitative Data Analysis
When is it performed?
After data collection complete
Simultaneously with data collection
(Constant comparative data analysis)
Data collection & analysis “staged
Two major styles:
Fracturing, grouping, gluing (bits)
Circling and parking (whole)
Fracturing, Grouping, Gluing
Most common style of qualitative analysis
Major analysis strategy is that of coding:
ie data are:
- fractured (divided into labelled bits=codes)
- categorised (codes are grouped into tentative categories & labelled)
- integrated (linked/glued together)
Circling & Parking
Data set treated as mass of information
Not fractured into small bits/sections
Aim: understanding overall themes
‘Circle’ data set
‘Park’ for closer scrutiny
‘Circle’, ‘Park’, ‘Circle’ etc until understanding complete
Other Styles of Analysing Qualitative Data
“Magnifying glass” style: data that are focus of study subjected to minute scrutiny
Layering & Comparing: eg in Ethnography; data analysed in layers
Quantitative Descriptive Statistics
Statistical procedures: give organisation & meaning to numerical data
Descriptive statistics: describe, organise, summarise raw data
of Inferential Statistics: make predictions & generalise findings
Two important functions of descriptive statistics
Organisation of data into figures eg piecharts, histograms, scatter plots, tables, line graphs
-Graphical & numerical techniques for organising & interpreting data
So enable trends & differences to be noted & calculation of simple statistics
Condense/reduce large quantities of numerical information into meaningful units
-Can be condensed & summarised using statistics eg measures of central tendency & measures of variability
Statistics:
Levels of Measurement
Nominal
Ordinal
Interval
Ratio
Statistics:
Ordinal Measures
Used to show relative ranking of events/objects
Conveys more information than nominal data
eg “Please rate the quality of nursing care received in this hospital”
Very poor, poor, average, good, very good, excellent
Statistics:
Frequency Distribution
The most basic way of organising data
The number of times each event occurs is counted Often expressed as percentages (%) May be numbers of cases (N or n) N=number in the whole sample n=number in subgroup of the sample
Statistics: Interval Measures (Continuous)
Differences between scores/measures can be treated as equal
There is a specific numerical distance between each of the levels
No absolute zero
eg Temperature (°C); blood pressure (mmHg)
Statistics:
Ratio Measures
Also continuous
Show ranking of events/objects with equal intervals
Do have absolute zero: makes the ratio of scale values meaningful
Highest level of measurement
eg height, weight, distance, pulse, blood pressure
Statistics:
Measures of Central Tendency
Single central score- enables summarising distribution of data set
Describe the centre of a distribution of scores
Three measures most common:
Mode
Median
Mean
Statistics:
Normal Distribution
Symmetrical & bell-shaped
Mean is at the centre
Most common score (mode) at the centre
Middle score (median) at the centre
Shows normal distribution- few values low or high- most in middle.
Statistics:
Skewness, Symmetry and Kurtosis
Skewness: measure of the asymmetry of the distribution of scores
Can have =ve/-ve skew
Symmetry: 2 halves of a distribution are mirror image of each other
Kurtosis: Related to the ‘peakness’ or flatness of a distribution
Statistics:
Measures of Variability/Dispersion
Concerned with the spread of data
Variability answers:
- Is the sample homogeneous or heterogeneous? - Are the samples similar or different?
Measures of variation describe extent to which individuals/scores in sample vary
Most common measures are:
- range - variance - standard deviation
Statistics:
Range
Simplest & most unstable measure of variability
The difference between the highest & lowest scores
Disadvantage: depends on the 2 extreme scores only (outliers)
Can use difference between other scores e.g. semi-quartile range
