issues with mixed methods research
DESCRIPTION
Group GG Asmah Karim 11M8140 Jenny Tan Feng Ling11M8134 Norul Faizah Hj Awg Jahri 11M8136 Rasidah Hj Mohamad 11M8139 Tan Soon Leong11M8073EE-5101-F. Issues with Mixed Methods Research. Content. Issues: Integration of paradigm Purpose Method of data collection - PowerPoint PPT PresentationTRANSCRIPT
Issues with
Mixed Methods Research
Group GGAsmah Karim 11M8140Jenny Tan Feng Ling 11M8134Norul Faizah Hj Awg Jahri 11M8136Rasidah Hj Mohamad 11M8139Tan Soon Leong 11M8073 EE-5101-F
Content
Issues: Integration of paradigm Purpose Method of data collection Data Analysis Validity/Legitimation Practicality Reporting Audience
Research Articles
Assumptions
Quantitative Qualitative
Ontological Reality is objective & singular
Reality is subjective & multiple
Epistemological
Researcher is independent from subjects
Researcher interacts with participants
Axiological Value-free & unbiased
Value-laden and biased
Rhetorical FormalImpersonal voiceQuantitative words
InformalPersonal voiceQualitative words
Methodological
Deductive processStatic design
Inductive processEmerging design
Integration of Paradigm
Creswell, 1994
Purpose
Often the purpose is not made clear.
Purpose necessitating mixed methods: Corroboration Complimentarity Development Expansion Initiation(Bazeley, 2002)
Method - Sampling
Quantitative QualitativeLarge, random Small, purposive
Statistical analysis HermeneuticStratified random or quota sampling
replaces purposive samplingThe inappropriate use of one method distorts, and potentially invalidates
the assumptions of another
Method - Triangulation
Initial conception To conduct parallel studies using
different methods to achieve the same purpose.
To provide corroborating evidence for the conclusion drawn.
Technique of validation
Used loosely as a synonym
Data Analysis – Quantitizing data
Meaning becomes fixed and single-dimensional.
Counts and proportions assume a level of scaling.
Potential disjucture between 0 and 1 in a continuous scale.
Data Analysis – Quantitizing data
Data from qualitative coding will be nominal or ordinal rather than interval
Distribution may be unknown and normality cannot be assumed.
Ignores the meaning of missing data/outliers.
Validity / Legitimation
Sample integration legitimation Making statistical generalization Meta-inference quality Population transferability
Practicality
Time and costly Knowledge of the multiple methods Their assumptions Analysis procedures Tools Ability to understand and interpret
results Disciplinary training
methodological prejudice
Reporting
Like writing qualitative research
Unlikely to follow traditional format
How best to present ideas and evidence? The degree to which quantitative and qualitative
components can or should be integrated.
A report which is disjointed and potentially repetitive.
Better to progressively unveil relevant evidence towards concluding, than to organize on the basis of method used.
Audience
Readers may be unfamiliar with either method.
The extent to which the consumers value the meta-inferences.
Research papers
Examples:
Where to post your Q?
Post your Q. in the assigned group
Q & A