23.07.2013 Views

embedding sustainability in organizational culture - Network for ...

embedding sustainability in organizational culture - Network for ...

embedding sustainability in organizational culture - Network for ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

appeNdIx Chapter 1: a: Introduction Methodology<br />

64<br />

How are they do<strong>in</strong>g it?: For each practice, we asked ourselves, how are they<br />

go<strong>in</strong>g about it? Initially, we used codes such as talk, model, system, rule, procedure<br />

and codify. We began to group these <strong>in</strong>to two categories of approaches: <strong>in</strong><strong>for</strong>mal and<br />

<strong>for</strong>mal.<br />

When are they do<strong>in</strong>g it?: Aga<strong>in</strong>, our aim was to assess where each organization<br />

was along the <strong>susta<strong>in</strong>ability</strong> trajectory (from just start<strong>in</strong>g out through to truly<br />

<strong>embedd<strong>in</strong>g</strong> <strong>susta<strong>in</strong>ability</strong>). Un<strong>for</strong>tunately, the data sources frequently did not lend<br />

themselves to this k<strong>in</strong>d of assessment. While we cont<strong>in</strong>ued to attempt to code the<br />

data <strong>in</strong> this way, <strong>in</strong> the end, we did not make use of these codes <strong>in</strong> our analysis.<br />

Our <strong>in</strong>ability to sort the practices <strong>in</strong> this way led to one of our recommendations<br />

<strong>for</strong> future research—that researchers attend to where an organization is along the<br />

<strong>susta<strong>in</strong>ability</strong> trajectory.<br />

As we proceeded with cod<strong>in</strong>g, we began to develop a tentative framework as a way<br />

to organize the practices. We observed that any given practice might be coded as:<br />

<strong>in</strong><strong>for</strong>mal/fulfillment; <strong>for</strong>mal/fulfillment; <strong>in</strong><strong>for</strong>mal/<strong>in</strong>novation; or <strong>for</strong>mal/<strong>in</strong>novation<br />

and that we were f<strong>in</strong>d<strong>in</strong>g practices with each of these characteristics. This led to<br />

the development of a framework that contrasts the goal of the practice with the<br />

approach to mak<strong>in</strong>g it happen. (More <strong>in</strong><strong>for</strong>mation on the development of this<br />

framework is provided below).<br />

As a f<strong>in</strong>al quality assurance check on our data set, at the end of our prelim<strong>in</strong>ary<br />

cod<strong>in</strong>g round, we sorted all of our sources by ascend<strong>in</strong>g number of practice codes.<br />

We reviewed all studies with zero or only one code to confirm their <strong>in</strong>clusion or<br />

exclusion from the study. We excluded sources without any practice codes or other<br />

relevant codes and documented our rationale <strong>for</strong> exclud<strong>in</strong>g them. These exclusions<br />

are reflected <strong>in</strong> the summary table presented above.<br />

critical appraisal of the data<br />

A key part of the systematic review process <strong>in</strong>volves mak<strong>in</strong>g a critical appraisal of<br />

the review f<strong>in</strong>d<strong>in</strong>gs. Our <strong>in</strong>itial desire was to be able to make an assessment of the<br />

strength of support of each research study. Our desire was to be able to compare<br />

the studies <strong>in</strong> terms of sample sizes and effect sizes. We did attempt to classify our<br />

studies <strong>in</strong> this manner. However, the vast majority of the studies <strong>in</strong>cluded <strong>in</strong> this<br />

review did not report effect sizes, reflect<strong>in</strong>g the tendency <strong>for</strong> research <strong>in</strong> this area<br />

to make use of qualitative case studies of ‘<strong>susta<strong>in</strong>ability</strong> leaders.’ In our view, this<br />

f<strong>in</strong>d<strong>in</strong>g alone presents a reflection on the state of the literature <strong>in</strong> this doma<strong>in</strong>.<br />

An additional issue that we faced was that many research studies made reference<br />

to the efficacy of particular practices that went beyond the empirical f<strong>in</strong>d<strong>in</strong>gs of the<br />

paper. This is not unique to this research doma<strong>in</strong>, but it did present challenges <strong>for</strong><br />

our method, given that our unit of analysis became the practices themselves, rather<br />

than a given study. Our response was to switch from assess<strong>in</strong>g the strength of each<br />

study and <strong>in</strong>stead, to code each <strong>in</strong>dividual practice <strong>in</strong> term of its degree of empirical<br />

support. We used three codes <strong>for</strong> this purpose: proposed but not yet tested (P);<br />

empirically supported (E); and empirically tested and unsupported (N).<br />

Practices were classified as proposed but not yet tested (P) if the practice was<br />

proposed <strong>in</strong> a practitioner report, a book aimed at practitioners or <strong>in</strong> the case where<br />

empirical papers proposed a practice that went beyond the empirical f<strong>in</strong>d<strong>in</strong>gs of that<br />

particular paper. Practices were classified as empirically supported (E) when the<br />

particular practice was the subject of empirical test<strong>in</strong>g <strong>in</strong> the article and the practice<br />

received empirical support. Practices were classified as empirically tested and<br />

unsupported (N), when the particular practice was the subject of empirical test<strong>in</strong>g <strong>in</strong><br />

the article and the practice did not receive empirical support. In the end, there were<br />

only three <strong>in</strong>stances of practices that were reported as unsupported out of a total<br />

pool of 1695 <strong>in</strong>stances of practices. In each case, it was not the entire practice that<br />

was unsupported, but rather an aspect of the practice. These <strong>in</strong>stances are addressed<br />

<strong>in</strong> the text <strong>in</strong> the discussion on the particular practice and are marked <strong>in</strong> Table A2.<br />

data synthesis<br />

Our cod<strong>in</strong>g process was systematic, <strong>in</strong>ductive, and emergent (Holton, 2007). Our<br />

codes evolved through constant iteration compar<strong>in</strong>g practice to practice to f<strong>in</strong>d<br />

patterns, and to establish similarities and differences between practices. The codes<br />

were assigned by go<strong>in</strong>g l<strong>in</strong>e by l<strong>in</strong>e through the data to identify and label actions,<br />

characteristics, and qualities as they arose. Based on an evolv<strong>in</strong>g framework that<br />

aimed to group practices accord<strong>in</strong>g to their approach and their goal, we began<br />

tentatively group<strong>in</strong>g practices <strong>in</strong>to categories. Codes that seemed to be similar were<br />

grouped <strong>in</strong>to tentative categories. Practices <strong>in</strong> a category were compared and sorted,<br />

and the properties of each category were articulated and ref<strong>in</strong>ed.<br />

The process of group<strong>in</strong>g, def<strong>in</strong><strong>in</strong>g and ref<strong>in</strong><strong>in</strong>g codes was achieved through an<br />

active and ongo<strong>in</strong>g discussion between the co-authors. We spent hundreds of hours<br />

review<strong>in</strong>g and discuss<strong>in</strong>g each and every <strong>in</strong>stance of a practice (1695 quotations <strong>in</strong><br />

total) and made decisions about where to place it, how it was similar to or different

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

Saved successfully!

Ooh no, something went wrong!