research for nurses

RESEARCH FOR NURSES : 

RESEARCH FOR NURSES Part of Advanced Midwifery Diploma 2010

What is research : 

What is research A systematic inquiry- planned, organized with diligence Uses disciplined methods & processes- purpose, actions, goal. Used to: develop, refine or expand & validate knowledge Researcher actively looking for information that is ‘new’ Precise, accurate with no bias.

What is nursing research? : 

What is nursing research? A systematic inquiry about issues important to nursing: practice, education, administration & informatics Increase nursing body of knowledge Improving & supporting practice -patient outcomes

Why do nursing research? : 

Why do nursing research? Best possible patient care that is evidence based Cost effective and efficient management Education of next generation is firmly based on evidence and fact. Generally so nursing decisions and actions based on evidence and not supposition/ inheritance In law courts an action taken by a nurse comes under scrutiny –is it research based?

Where does nursing information come from? : 

Where does nursing information come from? Inherited –”we have always done it this way” Specialized fields –their authority& information adapted to nursing Experience, trial & error learning Logic reasoning Generalisation from inductive specific knowledge “if it works there it should work here’ Research

NURSING RESEARCH v RESEARCH : 

NURSING RESEARCH v RESEARCH Nursing research/ quality improvement/ research can all be misconstrued. RESEARCH =philosophical underpinning, empirical base. Generalisable. New knowledge. Measurement of a phenomenon is possible and provides evidence for its existence. QUALITY IMPROVEMENT= pragmatic, useful and practical application of knowledge in natural settings and setting standards for them. Effectiveness of an intervention –it is EVALUATION.

NURSING RESEARCH v RESEARCH : 

NURSING RESEARCH v RESEARCH NURSING RESEARCH = not always generalisable, use non-experimental methods, ? Philosophical base. No risk to patients. Questions arising often cant be framed in research terms. EVIDENCE BASED PRACTICE =incorporation of evidence from research, clinical expertise and patient preferences into decisions about the healthcare plan for an individual

RESEARCH PROCESS : 

RESEARCH PROCESS PRECONCEPTUAL DISSEMINATION DESIGN & PLAN INFORMATION ANALYTICAL EMPIRICAL

RESEARCH PROCESS : 

RESEARCH PROCESS Phase 1 CONCEPTUAL PHASE: Identify the problem/ research question Determine purpose of study & significance Literature search Formulate research hypothesis Specify group to be studied Define theoretical framework

Research Process : 

Research Process Phase 2 : Design & planning Methods to be used for: Sampling techniques Data collection tools & instruments Data collection Validity/ reliability Ethics Piloting & revising

RESEARCH PROCESS : 

RESEARCH PROCESS Phase 3: the empirical phase Collecting data Preparing the data for analysis Phase 4: the analytic phase: Analysis of data Interpretation of results Phase 5: the dissemination phase Communicate findings by lectures/ publishing/ presentations etc Utilization of findings

RESEARCH UTILIZATION : 

RESEARCH UTILIZATION Using research from another discipline by translating the knowledge into real life based applications. RU can be indirect research utilization to change nurses thinking Direct RU is direct utilization of findings into patient care Persuasive utilization : persuading policy changers to make changes to nursing care based on utilizing findings from another disciplines research.

EVIDENCE BASED PRACTICE : 

EVIDENCE BASED PRACTICE Integration of best practice research evidence with clinical expertise and patient values. RU begins with research EBP begins with a clinical question and should offer solutions to improvement of health care. EBP provides an important framework for self directed learning Concerns are: exaggeration of claims, disregard of patient inputs and little attention to qualitative research to answer questions

BARRIERS : 

BARRIERS Management reluctance Cost constraints Time lag between research findings & dissemination of evidence Accessing research Presentation of findings Biases Limitations Fear from staff

THE RESEARCH PURPOSE ( focusing your research) : 

THE RESEARCH PURPOSE ( focusing your research) 1. Answer: Why are you conducting this research?: Identify problem Describe problem Explain problem Predict an outcome from a set of circumstances Evaluate a process/ product Develop further Compare Investigate Find a solution to Demonstrate a relationship to nursing

Where do you find research problems? : 

Where do you find research problems? Practice –querying why something is/ isnt done and why Education Management Ethics History

THE RESEARCH PROBLEM : 

THE RESEARCH PROBLEM A research problem is a perplexing or troubled condition . Problem statement: saying what the problem is and giving the argument for studying it Research questions: the specific queries about the research problem that the researcher wants answers to Hypothesis: an explanation based on the limited evidence to date as the starting point for further investigation. Statement of purpose: researchers summary of the overall goal of the study Research objectives : the specific aims the researchers hopes to achieve from the research

RESEARCH PROBLEMS; developing & refining them : 

RESEARCH PROBLEMS; developing & refining them IDEA of a problem Phenomenon you would like to validate or enquire about BRAINSTORM YOUR IDEA Identify the topic of interest LITERATURE REVIEW Refine problem, previous results etc FORMULATE A RESEARCH PROBLEM Is it appropriate/ feasible/ accurate? If not refine.

Appropriateness : 

Appropriateness Significance of your study to nursing? Will the evidence you gather potentially contribute meaningfully to nursing? Is it important? Is it worth doing? Will there be benefit to nurses/ patients/community? Will practice improve? Is the problem researchable? Ethical/ moral issues. Must yield facts Is it a feasible research problem? Time, resources, subjects, collaboration, facilities, equipment, ethical considerations Are you interested? You need a lot of drive, persistence, time energy to do a study – do you have it?

ELEMENTS FOR FORMULATING A RESARCH PROBLEM : 

ELEMENTS FOR FORMULATING A RESARCH PROBLEM WHAT? (Object) WHO? (Subject) HOW? (Actions) WHERE? (Context)

VARIABLES : 

VARIABLES INDEPENDENT VARIABLE: or treatment, explanatory, intervention, input variable. This is the ‘thing’ that is manipulated to see what its effects are on the dependent variable. WHAT DO I CHANGE? DEPENDENT VARIABLE: or outcome, effect variable. This variable responds to the independent variable. WHAT DO I OBSERVE? CONFOUNDING VARIABLE: or extraneous variable, a factor that could play a part in the study but which is not part of the study e.g. weather. May not be aware of it originally but find a factor influencing the study outside of independent/ dependent variables set out. WHAT OTHER FACTORS MAY CONTRIBUTE UNINTENTIONALLY? CONTROLLED VARIABLES: those factors that need to be kept controlled, the same to get the effects of the independent variable only. WHAT MUST I KEEP THE SAME?

THESE VARIABLES CAN BE FURTHER CATAGORIZED : 

THESE VARIABLES CAN BE FURTHER CATAGORIZED Discrete: a whole number and you can only belong in one category eg blood group, gender Continuous: any number or part of a number on a scale e.g. BP, weight Binary: only a yes or no answer Ordinal (scale): Likert scale type answers e.g. severity, degree of difference. Nominal (scale): groups have specific labels e.g. rocks are igneous, sandstone, sedimentary.

CATEGORIES OF VARIABLES : 

CATEGORIES OF VARIABLES NUMERICAL DISCRETE CONTINUOUS Blood type, gender, number of children etc. Height, Weight, blood pressure, age. CATEGORICAL BINARY ORDINAL NOMINAL Yes/No Severity of illness Marital status, Ethnic group

WHAT IS A RESEARCH HYPOTHESIS? : 

WHAT IS A RESEARCH HYPOTHESIS? Once you have identified your problem and done the literature search you can now make a formal statement on what you expect from the relationship between a set of variables. It’s a prediction of what the outcomes should/ shouldn’t be Used in quantitative research to direct experimental/ quasi-experimental studies or to test theories. The hypothesis is what is tested before it can become incorporated into a theory. Guides the investigation and provides a focus for the study.

A HYPOTHESIS MUST: : 

A HYPOTHESIS MUST: State a predicted relationship between a set of variables Be conceptually clear, specific and stated as simply as possible Be consistent with the existing body of research findings and with logical reasoning Be testable with available current techniques Relate to a matter which can be clearly defined empirically

TYPES OF HYPOTHESIS : 

TYPES OF HYPOTHESIS NON DIRECTIONAL v DIRECTIONAL SIMPLE v COMPLEX NULL v RESEARCH

RELATIONSHIPS : 

RELATIONSHIPS

RESEARCH DESIGN : 

RESEARCH DESIGN The plan of how the research will take place. The aim of the study The underpinning of the study –lit search. Context in which the study will be undertaken Techniques to be used to: identify the population, sampling, ethical considerations data collection methods etc must all be put into a plan. (THE ACTIONS)

QUALITATIVE APPROACH : 

QUALITATIVE APPROACH Qualitative approach: Looks at a phenomenon in its context and entirety Peoples lives, experiences No preconceived ideas No control over the context of the research Narrative information collecting No formal structured instruments used Social sciences use this

QUANTITATIVE APPROACH : 

QUANTITATIVE APPROACH Restrictive and highly controlled environment Preconceived idea of how concepts are interrelated Structured process of data collection Context controlled Focuses on objectivity Numeric data for statistical analysis

EXPERIMENTAL /QUASI_EXPERIMENTAL & NON_EXPERIMENTAL : 

EXPERIMENTAL /QUASI_EXPERIMENTAL & NON_EXPERIMENTAL Experimental: manipulates the independent variable, controls the situation and uses randomisation. Usually has control & experimental groups. Quasi-experimental: similar to experimental but some of the features cannot be manipulated or controlled to make it truly experimental. Non-experimental: manipulation of the independent variable is impossible or impractical and other research methods are impractical to use. Can have descriptive designs (retrospective, prospective longitudinal, cross-sectional) or correlational designs.

TYPES OF DESIGNS : 

TYPES OF DESIGNS Experimental designs: True experimental Quasi-experimental Non-experimental designs: Descriptive Correlational Traditional qualitative designs: Phenomenology Grounded theory Ethnography

EXTERNAL VALIDITY : 

EXTERNAL VALIDITY External validity: means the findings /cause-effect relationship from one study can be generalisable to the population or other conditions. High –low external validity. High = can be applied to target population + to other populations. Low= one cannot be certain that what you have found could apply to another group on the other side of the world. Small samples cause low external validity.

THREATS TO EXTERNAL VALIDITY : 

THREATS TO EXTERNAL VALIDITY Placebo/ Hawthorne/ reactive effect aptitude- treatment interaction eg the participants know they are being tested or that they have the correct drug. Situation: place, time etc all affect responses. Pre test effects- researcher effect Post test effects- researcher effect

INTERNAL VALDITY : 

INTERNAL VALDITY The causal relationship between variables is properly demonstrated satisfying: The cause preceeds the effect. Must not ‘predict’ the effect as this can influence your thoughts, design etc, The cause and effect are related There are no other plausible explanations

THREATS TO INTERNAL VALIDITY : 

THREATS TO INTERNAL VALIDITY Confounding variables. If the researcher can exclude confounding variables as influencing the study results then a rival hypothesis may be born Selection bias. The selection of participants eg attitudes, willingness, motivation may affect results may affect results even if some can be taken into account when sampling. History: participants may give one result now then another result later if e.g. a natural disaster occurred, political influences occurred Maturation: participants grow up and change attitudes. Affects longitudinal studies Repeated testing: the participants become biased. They know what to expect. Instrument change: the type of instrument used can affect results Mortality: participant may die/ exclude himself so altering demographics Experimenter bias –inadvertent effect on study.

POPULATION, SAMPLES &SAMPLING : 

POPULATION, SAMPLES &SAMPLING POPULATION: the total i.e. whole world, elements that fit the criteria for your research study. People, objects, substances e.g. blood samples TARGET POPULATION: the entire set of individuals that meet the criteria set out and the population to which the researcher will generalise the research findings to. ACCESSIBLE POPULATION: the target population may be too big to include as a whole so the researcher selects a population from the target that is accessible e.g. geographically. CRITERIA OR THE FRAMEWORK FOR INCLUSION INTO THE PROJECT MUST BE DEFINED EXPLICITLY.

SAMPLING : 

SAMPLING SAMPLING FRAME: a complete list from the population of sampling elements from which the sample will be selected. SAMPLING: the sample is selected by sampling methods a portion of the identified population that will represent the population. SAMPLE: The final set of people or elements from the population that will be included in the study. A representative sample means you have tried to keep your sample as true to the population as you can, replicating population variables as closely as possible. SAMPLING PLAN: the whole detailed process that will be used to select the sample from the target group including what sampling method will be used SAMPLING METHOD: the instrument, device that will be used to select the sample SAMPLING BIAS: the researcher ‘guides’ the sampling selection either unconsciously or consciously.

SAMPLING METHODS : 

SAMPLING METHODS PROBABILITY(RANDOM) SAMPLING: all elements in your population have an equal chance of being selected into the sample. More representative of the population. Best for quantitative or outcome research. Less opportunity for systematic bias. TECHNIQUES FOR RANDOM SAMPLING: simple, interval, stratified random, cluster random.

SIMPLE RANDOM SAMPLING. : 

SIMPLE RANDOM SAMPLING. Each of the criteria used to select the population initially are used and a random pick from each of the criteria is done. It is a one stage technique Each subject has an equal chance of being selected The study population can be identified and listed E.g putting names in a hat and drawing x number randomly. Using a table of random numbers or using computer generated selection of random numbers.

INTERVAL or SYSTEMATIC SAMPLING : 

INTERVAL or SYSTEMATIC SAMPLING As it says, a subject is selected simply because it is at the right interval originally selected e.g. every 10th subject on a list. The list must not be in a systematic order e.g. alphabetical. A sample size is determined before hand e.g. 30 subjects are needed from the list. A Random starting point is chosen. The interval is determined by a formula relating the size of the population to the size of the sample required from that population.

STRATIFIED RANDOM SAMPLING : 

STRATIFIED RANDOM SAMPLING When the researcher knows some of the elements in a population and in order to get a representative sample from each of these elements a selection must be made in each group. A subject can only belong to 1 layer (stratum) The groupings must correlate to the dependent variables being examined. A random number sampling method is then applied to each layer.

CLUSTER SAMPLING : 

CLUSTER SAMPLING Researcher starts with the largest population and systematically brings it down to the final target population from which a sample can be taken Used when the population is just too big or it would be too expensive, geographically too widespread. Or used when there are just too many criteria or unknown criteria to do an objective random process.

NON- PROBABILITY OR NON- RANDOM SAMPLING : 

NON- PROBABILITY OR NON- RANDOM SAMPLING This method does not necessarily represent the population Used when the population cannot be probability sampled accurately or the researcher doesn’t want probability sampling. Used for qualitative research Types of non-probability sampling: convenience, quota, judgemental, snowball

CONVENIENCE SAMPLING : 

CONVENIENCE SAMPLING Selecting a subject just because it arrives at the right time and place to be selected. Bias is strong this way Inexpensive and require less time.

QUOTA SAMPLING : 

QUOTA SAMPLING Equivalent to stratified where a specific group is targetted to get a quota of them. Usually used to make sure under represented sections are utilized. Decreases bias. If the demographics of the population says 60% men, 30% women and 10% children then accidental sampling can go for each stratum until the percentages are achieved then the gates close on that stratum.

SNOWBALL SAMPLING : 

SNOWBALL SAMPLING The researcher gets a couple of subjects then gets them to enlist a couple each and so on until a sample size is obtained. Good for targetting people who don’t want their identity known e.g. drug users.

JUDGMENTAL SAMPLING : 

JUDGMENTAL SAMPLING The researcher deliberately selects subjects to be included in the study. Qualitative research where you are not so much interested in an outcome but in the information you will gain. The researcher continues researching subjects until saturation of information occurs i.e. subjects are saying what you already know from previous interviews etc.

DATA COLLECTION : 

DATA COLLECTION FACT: an observation that can be verified time and again by empirical evidence MEASUREMENT: a unit of analysis e.g. cm meters, number of people. DATA: measurements that are collected in a systematic scientific way.

WHAT DATA? : 

WHAT DATA? This must be clarified and stipulated in the design stage. Exactly what data is sought, the level of measurement and the measurement scales to be used. TYPES OF SCALES: nominal, ordinal, interval, ratio

SCALES : 

SCALES NOMINAL: mutually exclusive catagories e.g. gender is mutually exclusive, you cant be a member of all the groups. ORDINAL: catagorised and ranked but cant have a numerical value e.g. pain can be slight, moderate severe but cant actually give it a quantitative value. RATIO: must be able to have zero. Time, length, weight.

Measurement Scales : 

Measurement Scales

HOW? : 

HOW? Must have detailed a data collection plan in the original designing of the study. What instrument will be used? Why was that specific instrument selected? Is it the correct one –will it lead to bias? Incorrect data?

WHO? : 

WHO? Researcher only? If others are going to assist they must be adequately briefed and educated on the tool and methods to be used with spot checks throughout the study to make sure the daa is valid otherwise the whole study is wrong.

WHERE? : 

WHERE? Seems simple but must be given thought. The site identified cold be problematic e.g. no toilets, no shelter, no privacy.

WHEN? : 

WHEN? Again important to give thought to this. It is pointless arriving at a site to find it is a public holiday and no one is there. Research the when to make sure it is acceptable to all Will this be a once off collection or a repeated collection? Pilot the study to find out problems.

VALIDITY OF THE INSTRUMENT BEING USED : 

VALIDITY OF THE INSTRUMENT BEING USED The instrument being used to collect data must be accurate, reliable and valid. CONTENT VALIDITY: how well does the instrument represent all the different components or variables it is meant to measure? CRITERION VALIDITY: using another valid instrument and measurements to compare yours to and see if it is accurate CONSTRUCT VALIDITY: what is the instrument set to measure? FACE VALIDITY: is this a reasonable way to gain data.

RELIABILITY : 

RELIABILITY Can the instrument be relied upon to yield consistent results even in another context? STABILITY: test, retest method EQUIVALENCE: can the same result be achieved with different data collectors. INTERNAL CONSISTENCY: do all items that make up the instrument measure the same variable. For 1 construct at a time