Own Research Flashcards
Research method includes :
an underpinning philosophy ( research paradigm), a strategy for recrutting participants, method, data collection, methods of data anaylsis, strategies for facilatating reliability and validity ( quantitve methods), or trustworthniess ( qualitative methods),
3 essential elements of a true experiment are :
Randomisation, Manipulation ( of the indepedent variable) and Control
Experimental design can be classified into 3 main groups
: Between-participants, Design within-participants design and Matched pairs design
Between- participants design
Most basic and commonly used
Involves 2 separate groups of participants
There are 3 between participants design commonly used
Between- participants: Post-test only design
Participants are randomised in either experiment or control group
The experiement group review the ‘new’ treatment or intervention ( independent variable)
Whilst the control group receivews the conventional treatment or intervention
The impact of both are then measured ( dependent variable)
Limatation of between participants : post-test only design
Limatation: account has not been taken of pre-intervention dependent variable data.
Limatation: Differences in the dependent variable may be caused by individual differences between the 2 groups of participants rahter than the indepdent variable
Reducing individual differences can be carried out by; indentifying specific, predetermined criteria for participant inclusion in the study and the involvement of a large number of participants
Between- participants: Pre test, post term design
It addresses the weakness of the post-test only design
It involves the collection of depedent variable data before the indepdent variable is introduced
Pre test data sometimes known as baseline data
Limatation of Between- participants: Pre test, post term design
Limatation: problem with individual difference between groups
Limatation: Pre test may impact the depedent variable
Between- participants: Solomon four design
It is a variable of the pre test, post test design
Ppts in this group have the pre test and recieve experiemental intervention
Ppts in this group have the pre test and recieve conventional intervention
Ppts in this group do not have the pre test and recieve experimental intervention
Ppts in this group do have the pre test and receive conventional intervention
Within- participants design
Overcomes the problem of individual differences
Each participants is exposed at some point to both the experimental and control intervention
Limatation of within participants - design
Limatation: order effect known as practice effect: this is when whatever that is done first effects the final findings of the study
Limatation: Participants may drop out after first phase of the study
Strength of Limatation of within participants - design
smaller number of participants required : advantage time & cost
Matched Pairs Design
Gold standard experimental design as it attempted to address the weakness of the between particpants within-participants designs
Ppts in the experimental and control groups are matched as closely as possible
One membranes of each pair is randomised to either experimental or control group
The closer ppts are matched the between to minimise individual differences between ppts in the experimental and control group
Limatation of matched pairs design
Limatation: Matching can be difficult time consuming and this make the design costly
Limatation: The research should be careful not to over match for irrelevant variables
Limatation: some participants have to be exclused from the study if an appropriate ‘match’ can be recruited
Limatation: Even if careful matching there may still be individual differences
Examples of non experimental designs
Longitudinal
Delphi
Cross sectional
Retrospective designs
Cross sectional design
Known as one-hit or one-shot design
Most common non experimental design
This design involves a representative cross-section of a defined population which means that researcg can investigate a large number of participants at different stages of development simulatenously
strength & limatation of cross sectional design
Strength: Economical, can use large samples, findings are avaliable promptly
Limatation: provides only snapshot of here and now, low response rates may impact on findings, extraneous variables may impact on findings, data can be superficial
Longitudinal design
Commonly used in surveys( referred to as panel designs) and cohort studies ( reffered to as trend or follow up designs)
It involves the repeated collection of data at pre-specified intervals from the same sample over a set or ongoing period of time
Used in healthcare to identify patterns and trends to measure changes to compare data of individuals or groups
strength & limatation of Longitudinal design
Limatation: costly and time consuming, requies high level of researcher, risk of ppts drop out, over times ppts can become aware of the variables researchers are interested in
Strength: identifies trends, changes and developments over time, provided detailed and extensive data,
Delphi design
The aim is to reach agreement ( consensus) about the topic and data are collected from participants during a number of rounds of data collection
1st round - participants are asked to response to broad key questions about the topic and their responses are collated
2nd round- participants consider more focused questions which are based on the collated responses from the 1st round.
Ppts can add, alter or retract information in 2nd round
strength & weakness of Delphi design
Limatation: time consuming, no agreement over the ideal panel size, no clear defintion of consensus, bias in the selection ppts
Strength: Economical, encourages ppts honesty, anoymits and confidentiality of ppts remained,
is Delphi design a recruitment
Reffered to as panel designs
Some see it as a method rather than a design
Recuriment of a panel
Retrospective designs
Reffered as ex post factor, after the fact or correlational design
Commonly used in case-control studies and historical research
Involves the collection of data retrospectively whereby the dependent variable9 effect) has already has been affected by the indepdendent variable (cause)
Aim of retrospective design
The aim is to link the present situation with what has happened in the past
Ppts with a particular condition are recurited ( the cases) and compared with participants who do not have the condition( control). Differences between the 2 groups give the research an indication of the likely cause of the conditon
Capturing retrospective data about patients - accessing clincial databases - costly
What is research
Research is a planned systematic search for information
The steps of the research process
Identifying a problem, question or hypothesis
Formulate the research question or hypothesis
Search and evaluate the literature
Devise the research design and methods
Test the propsed methodology by carrying out pilot studies
Collect data
Conduct data analysis
Draw conclusions and write a report
