Course materials for/by Peter L. Patrick. May contain copyright material used for educational purposes. Please respect copyright.

 

Sampling Design

By Peter L Patrick

For LG554, Univ of Essex

 

 

The task

In this assignment you will:

o   Devise a simple variationist research question,

o   Select a sociolinguistic variable, and

o   Design a sampling plan for it.

(You will not need to actually carry out this research – the exercise is intended to give you practice in planning such work.)

 

Length: As with all our assignments in LG554, you should write a report of 1500 words or more. I will not penalise reports of up to 3000 words, but if you can express yourself clearly in 1500, please do. Under-length (ie, under-1500 words) reports may suffer from lack of content and be marked accordingly.

 

The Research Question

You must begin with a research question (RQ), and a sociolinguistic variable that it focuses on. The RQ is always what drives sampling, but for this assignment it is not the most important part. You may choose a standard RQ, adopted from other sociolinguistic studies, to be studied now with new data that you are proposing to collect. It is the sampling design that this assignment requires you to focus on.

 

For example, you might ask one (just one!) of the RQs below:

o      What is the social distribution of sociolinguistic variable X across the population of speech community A?

o      Is sociolinguistic variable X undergoing a change in progress in speech community A?

o      Is sociolinguistic variable X used according to the same linguistic constraints in speech communities A and B (or, groups A and B in a single speech community)?

 

Make it a sensible RQ, but keep it short and simple. Your proposed study should be of a size appropriate to an MA dissertation, or at the largest, a PhD thesis – don't be too ambitious!

 

The Sociolinguistic Variable

Since your focus is on the sampling design, choose a simple, possibly well-known sociolinguistic variable to study. It may be one you wish to study later in the term, or in your own MA or PhD – that would be efficient and give some  continuity across your assignments, or in preparing and refining your MA/PhD proposal – but it does not have to be.

 

(Macaulay 2009 gives many examples of classic sociolinguistic variables; if you wish to use a British English phonological variable, Foulkes & Docherty give many examples of current ones, with details.)

 

Again, make this section short & simple. You will have the chance later this year to investigate a sociolinguistic variable in detail, or to propose a new one for study.

 

The Sampling Design

Devise  a sampling design for a variationist speech community study to fit the RQ and sociolinguistic variable above.

 

You need to sample both across different types of speakers (speaker variables), and different types/styles/contexts of speech. You should take into account at least the first 4 classic social (explanatory) variables below, whether or not you ultimately propose to sample across all of them (probably not!). That is, if you choose not to vary your sample by gender or ethnicity, your discussion will explain the considerations involved in coming to your decision.

o      Style

o      Age

o      Gender

o      Social Class or Status

o      Ethnicity

o      Other relevant social factors, e.g. rural/urban identity, religion, sexuality, native-speakerhood, multilingual repertoire, etc.

 

(Note on Social Networks:

As we will see when we cover social networks, they are rarely empirically used by sociolinguists as an explanatory variable in themselves, and more commonly as a device in data collection and sampling, or invoked as an after-the-fact explanatory device without actual measurement. It is possible to treat social network as an explanatory aspect of social structure, but you will need a good understanding of the topic in order to do so; at this stage of the term you may not be ready.)

 

Details of Sampling Design

This is the longest section of your report, and should take up at least 2/3 to 3/4 of it. The Sampling Design should answer the questions:

o      How many speakers do you plan to collect data from?

o      Which social or sociolinguistic categories and factors do you plan to sample for, and why are they important to your RQ?

o      Which social or sociolinguistic categories and factors do you NOT plan to sample for, and why have you decided they could be excluded?

o      Discuss briefly any key points from the literature which bear on your decisions.

 

In addition, please specify:

o      What types of speech, reading and/or written data do you plan to collect from speakers?

o      How does each type of data bear on your RQ, and how will you collect it?

o      How much data of each type will you need?:

o   If interview or naturally-occurring speech, in minutes or hours;

o   If word-list, reading passage, etc., describe length.

o      How long do you think is likely to be needed to collect a complete set of data from each person in your sample? With that in mind,

o   How many days' work is the entire sampling process likely to take?

 

In your discussion, keep in mind concerns of validity and reliability as these may be interpreted in sociolinguistic survey work.

o   Does your design measure what you think, and claim, to be measuring? (=Validity)

o   If someone else used the same methods on a different group of speakers, or comparable methods next year on the same group, would their results be highly similar to yours? (=Reliability)

 

Finally,

o   What do you think might be some practical difficulties in completing this design? What steps could you take to mitigate or overcome those?

 

NB: You may find it challenging to cover all aspects in 1500-3000 words. Keep in mind that research grant applications also give you only limited space for a similar task. You'll probably want to write it out at length first, and then edit and condense it into the final version. Not every question or decision requires a paragraph of explanation or discussion – reserve it for those that need it most, in your judgment.

 

Class References relevant to this assignment include:

§  General: Macaulay (2009, chaps. 1, 3, 6); Milroy & Gordon (2003, chaps. 2 and 3); Schilling-Estes (2007, SocVar Chap. 9); Llamas (2007, RCS Chap. 2); Arksey & Knight (1999, Chaps. 1, opt. 4).

§  Age: Eckert 1997; Llamas (2007, RCS Chap. 8).

§  Ethnicity: Wolfram (2007, RCS Chap 9); Fought (2002, HLVC).

§  Gender: Cheshire (2002, HLVC); Coates (2007, RCS Chap 7).

§  Social Class/Status: Ash (2002, HLVC); Kerswill (2007, RCS Chap. 6).

§  Style: Schilling-Estes (2002, in HLVC); Bell (1997); Milroy & Gordon (2003, chap. 8); Macaulay (2009, chap. 4).

 

Details of Assessment for LG554

LG554 Sociolinguistic Methods page

References for Lg554

Peter L. Patrick's home page

Last updated10 November 2011