Research Designs Sebastian M Rasinger Quantitative Research in
Research Designs Sebastian M. Rasinger Quantitative Research in Linguistics. 2 e 2013 London: Bloomsbury S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
The research process Research Area Research Question S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Quantitative data • • can be put into numbers can be used for statistic analysis NOT necessarily ‘much data’ Deductive: theory hypotheses data confirm/reject H • Examples of quantitative variables – people’s age – test scores S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Qualitative data • • ‘words’, ‘text’, ‘qualities’, ‘patterns’ how is something, rather than how much interpretative inductive: theory is derived from the research results S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Research Designs • Longitudinal vs. cross-sectional – Change vs. status quo • Experimental vs. natural – Deliberate manipulation of variable/s? – (vs. quasi-experimental) S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Data – some practicalities • How much data? – As much as you need to make a coherent argument… – No need to prove beyond reasonable doubt – Does your data ‘make sense’ given the topic and previous theory S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Data – some practicalities (2) • Whatever method you use: pilot it!! – What’s going well? – What’s going horribly wrong? – Why have things gone wrong? Is it your instrument? Is it you? Is it both? – Adjust accordingly and, ideally, pilot again S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Causality • A causes B • Notoriously difficult to prove!! 1. Variables A and B must correlate with each other, that is, their values must co-occur in a particular pattern: for example, the older a speaker, the more dialect features you find in their speech. 2. There must be a temporal relationship between the two variables A and B, that is, B must occur after A. 3. The relationship between A and B must not disappear when controlled for a third variable. S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Reliability and Validity • Reliability: Does you method repeatedly measure what it’s supposed to measure? I. e. if you use the method under exactly the same circumstances, do you get the same result? – Split-half – Test-retest • Validity – Does you method measure what it’s supposed to measure? S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
Samples and Sampling • Part of population (P). Population = all people who show certain characteristics (e. g. year 2 undergraduate students at British universities) • Reflection of P, i. e. characteristics equal those of P • Ideally: random sample. Every member of P has equal chance to be member of sample • Realistically: opportunistic sample. Take what you can get. But check that it’s not completely off P use personal judgement S. M. Rasinger. 2013. Quantitative Research in Linguistics. 2 e. Bloomsbury.
- Slides: 12