Women's Decision-Making And Factors Affecting Their Choice Of ...

Women's Decision-Making And Factors Affecting Their Choice Of ... Women's Decision-Making And Factors Affecting Their Choice Of ...

epubs.surrey.ac.uk
from epubs.surrey.ac.uk More from this publisher
24.03.2013 Views

Hammersley 1992a, Hammersley 1992b; Morse 1998; Sandelowslci 1993; Silverman 1985). Sandelowsld (1993) suggests that participants may forget what they said and therefore are not in a position to verify the accuracy of researcher's accounts. They may also change their mind about what they gave in an interview afterwards. Another argument is that results of a study should be given to participants for their information, use and application, and not for verification (Morse 1998). Morse further argues that good qualitative research involves processes of synthesis, conceptualisation and abstraction. It involves extensive knowledge of the topic, the setting, and social science literature. The researcher uses this education to identify and create new knowledge. Participants, on the other hand, who are usually lay people, do not have these abilities. Additionally, research results are a synthesis of multiple participants' views which one participant may not understand. Morse (1998) considers verification by participants an extraordinary and unrealistic expectation. Furthermore, it is argued that participants cannot be relied upon to read the draft of the analysis with the same kind of critical spirit necessary for the task to be carried out successfully (Emerson 1988). Finally, Bloor (1983) suggests that member checks are limited to asking members to judge whether a researcher's account represents a legitimate elaboration and systematisation of the member's account. Another method suggested for improving the trustworthiness of a research investigation is triangulation (Hammersley 1992a; LeCompte and Goetz 1982; Lincoln and Guba 1985; Marshall and Rossman 1989; Sandelowski 1986). Triangulation refers to employing more than one method in data collection and analysis. Four types of triangulation have been identified, they are method, data, investigator, and theoretical (Campbell and Fiske 1959; Denzin 1970; Goodwin and Goodwin 1984b; Webb, Campbell, Schwartz et al 1981). 81

In method triangulation, different methods are used to address the same problem, in data triangulation, different data sources are used, in investigator, different investigators are used, while in theoretical triangulation different theoretical models are used. The aim of triangulation is to establish the convergent validity of findings from complementary approaches. It is argued that triangulation balances distorting effects of any single approach and permits data collected in one way to be used to check the accuracy of data collected in another (LeCompte and Preissle 1993). Some authors agree on the use of multiple methods in qualitative research, especially to extend the comprehensiveness of findings. Murphy et al (1998) argue that the use of multiple methods may encourage researchers to pay attention to the different perspectives that may be held, and the ways in which results are inevitably the product of the context in which they were produced. Additionally, it is thought that when data from two sources or methods produce dissimilar results, the researcher is prompted to consider how the differences came about, which enriches the analysis (Jick 1979). Similarly, it is argued that differences between two sets of data are just as important as the similarities (Hammersley and Atkinson 1995), and when the use of multiple methods uncovers discrepancies, they call for more investigation, which may lead to deeper understanding (Bryman 1988; Janesick 1994). Furthermore, multiple methods are thought to uncover some unique variance that may otherwise have been neglected by a single method (Oiler 1993). Although generally critical of triangulation, Silverman (1993) acknowledges the usefulness of multiple methods in overcoming the incompleteness of data drawn from a single source. On the other hand, some authors have warned against using triangulation as an end in itself, or looking at it as an inherent good (Knafl and Breitmayer 82

In method triangulation, different methods are used to address the same<br />

problem, in data triangulation, different data sources are used, in investigator,<br />

different investigators are used, while in theoretical triangulation different<br />

theoretical models are used. The aim of triangulation is to establish the<br />

convergent validity of findings from complementary approaches. It is argued that<br />

triangulation balances distorting effects of any single approach and permits data<br />

collected in one way to be used to check the accuracy of data collected in another<br />

(LeCompte and Preissle 1993).<br />

Some authors agree on the use of multiple methods in qualitative research,<br />

especially to extend the comprehensiveness of findings. Murphy et al (1998)<br />

argue that the use of multiple methods may encourage researchers to pay attention<br />

to the different perspectives that may be held, and the ways in which results are<br />

inevitably the product of the context in which they were produced. Additionally,<br />

it is thought that when data from two sources or methods produce dissimilar<br />

results, the researcher is prompted to consider how the differences came about,<br />

which enriches the analysis (Jick 1979). Similarly, it is argued that differences<br />

between two sets of data are just as important as the similarities (Hammersley and<br />

Atkinson 1995), and when the use of multiple methods uncovers discrepancies,<br />

they call for more investigation, which may lead to deeper understanding (Bryman<br />

1988; Janesick 1994). Furthermore, multiple methods are thought to uncover<br />

some unique variance that may otherwise have been neglected by a single method<br />

(Oiler 1993). Although generally critical of triangulation, Silverman (1993)<br />

acknowledges the usefulness of multiple methods in overcoming the<br />

incompleteness of data drawn from a single source.<br />

On the other hand, some authors have warned against using triangulation<br />

as an end in itself, or looking at it as an inherent good (Knafl and Breitmayer<br />

82

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!