Learner Resource 1
Research Methods Cards for Language Investigation
Naturalistic observation
Content analysis
Questionnaires
Means: recording real-life speech in a
normal environment, then transcribing it for
analysis. If you study a group you are part
of or very close to, this makes your work an
ethnographic study.
Benefits: primary data that is rich, exciting
and original, showing real language choices
being made on the spot. In-depth discourse
analysis can be carried out on the
transcripts. Data should be valid and
reliable.
Potential problems:
Observer’s Paradox – the very fact that you
are there recording something changes it.
Practical challenges – getting a clear
recording, finding time to transcribe,
knowing what to record and what to leave
out.
Obtaining informed consent – all
participants must consent to their speech
being recorded and analysed.
Means: taking written data that already
exists, or transcripts of planned speech, and
analysing its content in depth.
Benefits: less investigator bias and no
observer’s paradox, as the data existed
already. Data is authentic as it comes from
the real world.
Potential problems:
Context – with secondary data, you were
not there when the texts were created so
may have to guess contextual factors.
Choice of data – although you are not
affecting the language choices within the
text, bias can be introduced in the way you
select texts to analyse. It is essential to find
texts which are (a) directly comparable,
and (b) suitable for testing your hypothesis,
and (c) representative.
If texts were not already in the public
domain, or you do not have the right to
copy them for fair use and comment, you
must get informed consent from the
people who created them.
Means: asking participants questions about
their language use, views on language or
responses to mini-texts.
Benefits: enables you to find out attitudes
to language in a controlled way that is
relatively easy to analyse. Provides
quantitative (numerical) data.
Potential problems:
Questions need to be trialled first in a pilot
study to check they are easy to answer and
provide the intended types of answer.
Finding participants who are representative
of the type of people you want to investigate
and collecting responses can take longer
than you think.
Responses are limited by the format of the
questionnaire and participants’ willingness
to write long answers. This means the data
can be lacking in depth and interest.
Version 1
Setting up the Independent Language Research Project
1
© OCR 2015
Interviews
Experiment
Means: setting up a situation where you
can compare two or more groups of
speakers to see what is different about the
way they respond. Everything should be the
same for both groups except the variable
you are testing.
Benefits: enables you to clearly compare
males/females, older or younger speakers
or any other linguistic variable. Data is
relatively predictable. Fun to set up and
carry out.
Potential problems:
1. Validity of data – since an experiment
is automatically an artificial
environment, you are not recording
natural language behaviour.
2. Practical problems – an experiment
requires careful preparation. You need
to carry out a pilot study to check the
procedure works.
Version 1
Setting up the Independent Language Research Project
Means: talking to participants, using preplanned questions to elicit particular types of
language, or responses(missing comma here)
about their attitudes. Interviews need to be
recorded and transcribed, though the
transcripts can be selective if necessary.
Benefits: rich, spoken data, which is partially
controlled but still has an element of
authenticity. Most participants will stop
‘monitoring’ their speech if they get involved
enough in the conversation so validity will
improve.
Potential problems:
1. Recruiting participants who are willing to
give up their time to be interviewed.
2. Questions need to be planned in advance,
and decisions made about whether to
deviate from them if the conversation
moves in unexpected directions.
3. Participants need to be chosen to be
representative of the type of person you
are investigating.
4. Confidentiality must be maintained by
keeping participants anonymous.
2
Corpus analysis
Means: collecting a ‘body’ (corpus) of data
which represents the type of language you
are interested in and analysing it
quantitatively to find patterns. Qualitative
analysis can be done on selected
examples from the corpus.
Benefits: using a large amount of preexisting data enables you to test
hypotheses effectively. The data is reliable
because there is a lot of it.
Potential problems:
1. Getting access to a corpus of
appropriate data, and then managing
the analysis of it within the time limits of
an A-level project.
2. Context – with secondary data, you are
not aware of the contextual factors
affecting the language used.
© OCR 2015
OCR Resources: the small print
OCR’s resources are provided to support the teaching of OCR specifications, but in no way constitute an endorsed teaching method that is required by the Board, and the decision to
use them lies with the individual teacher. Whilst every effort is made to ensure the accuracy of the content, OCR cannot be held responsible for any errors or omissions within these
resources.
© OCR 2015 - This resource may be freely copied and distributed, as long as the OCR logo and this message remain intact and OCR is acknowledged as the originator of this work.
OCR acknowledges the use of the following content:
Please get in touch if you want to discuss the accessibility of resources we offer to support delivery of our qualifications: resources.feedback@ocr.org.uk
Version 1
Setting up the Independent Language Research Project
3
© OCR 2015