Microsoft Word doctoral thesis final[1]. doc

participated in the survey (15 from South Africa and 8 from the UK).  
 
b)  
Data gathering instruments and data gathering 
 
A structured sixteen page questionnaire consisting of a combination of close-ended pre-
coded items and open-ended items was used to obtain data from the 23 female academic 
HoDs from six universities in South Africa and the UK. The questionnaire was mailed by 
post to all the respondents (henceforth referred to as participants). After four weeks, a 
follow up letter together with a copy of the questionnaire was sent to all who had not 
responded. 
  
c)  
Analysis of quantitative data 
 
The SAS/STAT statistical package, version 9.1 was used to analyse the data captured 
from the questionnaire responses. Analyses undertaken included one-way frequencies; 
combined frequency tables; means calculations (standard deviations, minimum and 
maximum values) and cross-tabulations (details in chapter 4). 
 
1.6.2.2   Phase Two: the interviews 
 
a)   
Selection of informants 
 
Selection of informants for the interviews followed exactly the same procedure as for the 
survey. Informants from South African universities were first located through telephone 
contact with a human resources department person at the relevant institution. Once 
contact with the informant was established, a brief telephone interview was conducted in 
order to screen the prospective informant.  If she was found suitable, she was invited to 
participate in these focus group interviews. Six female heads of academic departments 
from the selected South African universities participated in the interviews.   

---

 
12
 
To locate suitable UK participants, the websites of the two selected universities were 
visited.  A thorough search of each university’s departments was done to locate female 
HoDs.   Communication with all the UK informants was by email. Three women heads of 
academic departments participated in the interviews. 
 
b)   
Data gathering 
 
The video conference focus group interview (V Conf-FGI) was used to gather data from 
nine informants (henceforth referred to as participants). Four interview sessions were 
held and participants attended according to pre-arranged schedules. Participants were 
emailed semi-structured interview guides prior to attending the interviews. The 
interviews were guided by the interview schedule and so were mostly “topic oriented” to 
borrow Lee’s and Fielding’s words (2004:533). Interviews were recorded on videotape 
and notes were taken at each interview session. These were examined for key themes and 
written out according to topic and filed. The videotape was played back to listen for, and 
to note, any patterns in the interview procedure and discussion which could be 
incorporated into subsequent sessions. 
 
c)   
Analysis of qualitative data 
 
Analysis of the data began as soon as the first interview was completed and continued 
throughout data gathering. At the end of the interviews the video tapes were transcribed 
and the data examined for key issues raised by participants in response to each topic. 
From these key issues, patterns were noted and data were categorised and discussed 
accordingly. Qualitative data from the interviews which corresponded with that from the 
survey were integrated. Patterns emerging from the data were noted and the findings were 
interpreted.  
 
 

---

 
13
1.6.2.3 Issues during research 
 
One of the main issues during research concerned the logistics of organising and 
synchronising the video conference focus group interviews.  The cost involved in 
conducting the interviews was another problem, particularly because of the different 
currencies and prevailing exchange rates between the UK Pound Sterling and the South 
African Rand. 
 
1.6.2.4 Trustworthiness of data 
 
In this study, both samples of participants (the survey and the interview) were gathered 
from a similar pool of respondents. This ensured a greater likelihood of obtaining reliable 
data. Moreover, standard conditions were established for each method of data collection.  
The equivalent questionnaire was mailed out to participants after being piloted and 
refined with the help of experts. For the interviews, the same procedure was followed. 
The interview schedule was piloted and refined before final administration. The 
participants were interviewed under similar conditions; that included starting time and 
duration of interview, interview schedule and role of researcher. To ensure reliability and 
validity of the data, the interviews were video-recorded, and additional notes were taken. 
The interviews took place under professional conditions. Every effort was made by the 
researcher to establish rapport with the participants.  
 
1.6.2.5   Limitations of research 
 
An acknowledgement of the limitations of this study may be in order. 
 
Firstly a random sampling technique was not employed as its aim was to focus on 
universities with women heads of academic department. Therefore, universities with no 
female heads of academic department were excluded from the study.  Consequently the 
research has a restricted dataset which limits generalisation of its findings. However, 
since the study is exploratory in nature and largely descriptive, no attempt was made to 
put forward hypotheses to be rejected or confirmed. Rather, an attempt was made to 

---

 
14
establish patterns and trends and relationships between certain variables in the 
quantitative part of the study, and to understand and describe the management and 
leadership experiences of women HoDs in the qualitative part.  
 
Secondly, although the sample was representative in terms of race (black and white), it 
was predominantly white and the results may thus reflect the experiences of this majority. 
Lastly, although the participants were drawn from a pool with similar characteristics, 
there were two groups, one for the survey and another for interviews. The reason for this 
choice was based on logistical considerations. Two countries halfway across the world 
from each other were involved. They are in different time zones. Calendars and heavy 
work loads were considered. The amount of time for completing the questionnaire and for