Chapter 5 – Methodology

Introduction

This chapter outlines the primary research methods used in the composition of this Thesis. It This explains the investigation methods used in the quantitative analysis by studying the relationship with one set of facts to another. Measurements of attitude using Statistical Package for the Social Sciences (SPSS), will produce quantifiable and generalisable conclusions. During the qualitative analysis, I will be more concerned with understanding individual’s perceptions of the Air Corps, relative to the questionnaire supported with evidence from previous chapters of the study. Where possible, primary information examined during the formative research period, has been relied upon. Secondary research information, has been used in the study in a selected, controlled way.

To the best of my knowledge this study has not been attempted before therefore, the conclusions reached in chapter six to eight are legitimate.

The tabulated results emanating from the attitudinal survey, constructs, a characterisation of the Air Corps and the respondents, who work there. It will be possible, to show, from the sample population selected, how respondents replied to preset data laterally, i.e. unit to unit, and vertically, i.e. according to rank throughout the organisation.

Research Question (Hypothesis)

Identify and describe the current management technique used in the Air Corps. Do requirements exist for achieving a high level of organisational management congruence in terms of the totality of organisational change?

Essentially the hypothesis contains one statement and poses two questions which will be addressed during the qualitative analysis of this study from chapter one to six. Areas of particular interest but not necessarily specific to the hypothesis or requiring further research at a higher doctoral level will be highlighted.

Questionnaire Data

The questionnaire method of eliciting data from the sample population selected, was decided, after, considering the size of the total available population. One principle disadvantage of using the survey questionnaire, is the difficulty of getting in depth detailed information which would more fully express the views of the respondents. Considering the size of the eligible population, and the difficulty of getting views representative of the whole population, the questionnaire survey instrument in this instance was considered the most suitable method. My aim, is to analyse responses, to a set of preset questions and relate the outcome to the posed research question.

A definite advantage of selecting this method of survey research, is that, it is personal thus,respondents will be more inclined to give more information once their identity is not revealed.

Only Enlisted Personnel are targeted in this study for several reasons. To include an analytical account of the Air Corps Officer Corps, the task would be enormous in terms of time and resources. Information emanating from such a study, would, be useful from an academic perspective. I feel confident, that, marginalising this group of organisational members from the current research will not in any way detract from findings or conclusions reached. In the interests of pursuing future research the Air Corps Officer Corps would be an interesting group to examine.

This study will show, that, two distinct remote groups co-exit in an organisational environment searching for identity, renewal and change. The groupings are Commissioned Officer’s and Enlisted Personnel. Fear, suspicion, manipulation, perceptions etc. are some of the misaligned processes at the core of the problem. This under use of organisational members is causing tension, and is preventing an acceptable level of congruence between the group’s from been reached.

The Sample

The population of Enlisted Personnel in the Air Corps, based, at Baldonnel, on the 1st of August 1996 is 848. This figure does not include 9 Air Corps Cadets (trainee officers) who are classified as Enlisted Personnel. Air Corps Cadets will not be a feature of this study and are therefore eliminated from the statistical computations.

A total of seven rank grades exist within the enlisted personnel ranking structure, and is listed below in ascending order of promotional advancement.

      1. Airperson Two Star (*)
      2. Airperson Three Star (*)
      3. Corporal (1st NCO rank grade)
      4. Sergeant
      5. Flight Quartermaster Sergeant
      6. Flight Sergeant
      7. Sergeant Major (Highest NCO rank grade)

The hierarchy of rank structure is stratified and clustered into Senior NCO’s, Flight Quartermaster Sergeant, Flight Sergeant, Sergeant Major and Junior NCO’s, Corporal and Sergeant. The analysis will show, if, any similarities or differences of views expressed by Senior and Junior NCO’s exist in the organisation down to unit level.

The Sample

It was decided to target 500 possible respondents, out of a total number of 848, or 58.96%, of the population was considered eligible to be targeted across the stratified ranking structure.

An up to date nominal role was obtained from Central A of all unit personnel in Baldonnel. The sample population was selected at random by systematically selecting every first person on each nominal role and there after every second person.

In deciding the sample size, the following formula was applied.1

INSERT FORMULA

A margin or error of 4% was selected, which is very low for such a study. If the respondents from the survey sample answered ‘yes’ for 50% of choices posed to preset data, then, I would be 96% confident that the yes percentage of the universe, in this case all Enlisted Personnel based at Baldonnel, had I studied the universe, would be between 46% and 54% or 4% either way of 50% Baldonnel, had I studied the universe, would be between 46% and 54% or 4% either way of 50%

Due to the small number of Sergeant Major’s (Sgt Maj’s), Battalion Quartermaster Sergeants (B/Q’s), Flight Sergeants and Flight Quartermaster Sergeants relative to the overall total figure of 848, it was decided to target all of the possible respondents in this stratified clustered rank category, to ensure that the returns would be definitely representative. All Enlisted Females of the Air Corps based at Baldonnel were also targeted due to their small numbers. The spread of possible targeting relative to the 50% sample size in the following rank grades is profiled. A total of 5 Sergeant Majors, 3 Barrack Quartermaster Sergeants, 53 Flight Sergeants, 14 flight Quartermaster Sergeants, 60 Sergeants, 96 Corporals and 269 Airpersons are considered eligible to be included in the target sample.

The possible outcome of the study, will be balanced, using the above method of selecting the population for study.

TOTAL POPULATION

Sgt MajBQF/SgtFQSgtCplArmnTotal%
5
0.59%
3
0.35%
53
6.25%
14
1.65%
120
14.15%
192
22.64%
461
54.36%
84899.99
5
100%
3
100%
53
100%
14
100%
60
50%
96
50%
269
58.35%
500
58.96%

Figure 5.1

Figure 5.1, shows the actual number of the sample population, according to rank. The total targeted sample according to rank, is also shown. A relative percentage figure within the groups, is also included. A total of 500 possible respondents was selected from 848 organisational members. Of this random stratified sample 58.96% or 500 comprise the eligible sample population.. Figure 5.2, shows the number of possible respondents in the universe according to rank at their various work locations in Baldonnel. Figure 5.3 shows the number of targeted eligible personnel.

The Survey Instrument

An attitudinal survey instrument, administered by ticking the appropriate box, totalling 128 questions was designed. The survey instrument was divided into thirteen sections as follows:

TOTAL TARGETED POPULATION ACCORDING TO UNITS

GP
HQ
Admin
Wing
No 1
Sp
Wing
No 3
Sp
Wing
Trg
Wing
Air SpTotal
Coy
Sigs
5th
Maint
Eng
Wing
Total
Sgt
Maj
210110005
BQ020100003
F/Sgt101811860953
FQ1423111114
Sgt8661010711260
Cpl491317111812396
Armn155733494119451269
Total4080639551511296500

Figure 5.2

TOTAL POPULATION ACCORDING TO UNITS

GP
HQ
Admin
Wing
No 1
Sp
Wing
No 3
Sp
Wing
Trg
Wing
Air SpTotal
Coy
Sigs
5th
Maint
Eng
Wing
Total
Sgt
Maj
210110005
BQ020100003
F/Sgt101811860953
FQ1423111114
Sgt161211221914125120
Cpl81826342235247192
Armn209656877138488461
Total571341041621239413170848

Figure 5.3

General Information

      1. Career Potential
      2. Attitudes to work
      3. Organisational communication
      4. Trust and motivation
      5. Decision-making Welfare
      6. Management
      7. Organisational culture
      8. Bureaucracy
      9. Power
      10. Job-design
      11. Organisation

Question one to twelve consists of ordinal data, i.e. age, marital status, educational standard etc. and question thirteen to one hundred and twenty eight consists of interval data, i.e. data which can be quantified and measured against a summated-rated scale such as the Likert Scale of measurement.

Within the questionnaire, a total of twelve open-ended questions was included. This was considered necessary, to give the respondent the opportunity to offer his/her own response rather that selecting an item from the attitudinal scale of measurement.

Many respondents gave interesting comments in response to the open-ended questions posed.

The Likert scale of measurement was selected for use with the survey instrument. This scale ranges from strongly disagree (SD) to strongly agree (SA) across five intervals. It is simple in construction and use. The midpoint of the scale undecided, is considered a neutral position for the respondent. The scoring of the of the five-step scale may go from left to right, i.e. SA (1), A (2), U (3), D (4), SD (5), or vice versa, i.e. 5,4,3,2,1 taking care that the ‘1’ response is consistently assigned to either the positive or the negative extreme. In this study I have assigned the ‘1’ to represent Strongly Agree.

Respondents are requested to place a tick inside the box which represents their strongest feeling to the question posed.

Instrument Design

Many considerations were reflected upon in relation to the design of the survey instrument.

The construction of the questionnaire proved to be difficult, and I was fully aware of the importance of avoiding adding personal bias to the question formations. As a Senior Non Commissioned Officer of the Air Corps (Flight Sergeant), I contemplated that my purpose for carrying out the survey might raise suspicion among superiors and subordinates of the organisation. To inhibit any possible threat of raising fear among my colleagues I expressed on the covering letter of each questionnaire that all responses from respondents will be treated in total confidence and in accordance of ethics of research methodology. The distribution of the survey instrument was sanctioned by the Base Commander, on 9th September 1996.

To-date Enlisted Personnel of the Air Corps have had little or no exposure to approved internal questionnaires. It was important to fashion the survey instrument in a balanced way to the number of questions posed, so that it, would be interesting, appealing and above all ‘user friendly’ which would help to increase the general response rate. To complete the questionnaire took approximately forty minutes.

Reasons for using SPSS

Considering the size of the target sample of 500 possible respondents, and the number of questions posed on the survey instrument totalling 128, I concluded that SPSS would greatly reduce time spent manually analysing data. Coupled with this, was the large number of tests which can be accurately performed using SPSS.

Statistical Package For The Social Sciences (SPSS), 6.1.3 version

Early versions of SPSS date back to 1968. The package is well known among university lecturers, undergraduates and post graduates. Health boards and educational research institutes are among the many bodies using the system.

Data Analysis

Preparing to use SPSS to analyse data is really a two-step process. The first step is often called data definition which involves establishing a computer file with names for variables, designated places for their storage. This is normally completed by using the ‘Add Variable’ and ‘Define Variable’ commands in SPSS for windows.

The second step is editing and coding the data. Verbal and written responses must be transformed into a number code for processing with SPSS.

Coding Data

Before the data can be entered into a file, the questionnaire needs to be edited. Each question on the questionnaire must be given a number in the identification field. This is necessary to avoid coding process errors. For example, question one, in the ‘New Data Field Window’ is transformed so that the variable cell is defined to represent question one (Q1), across the top of the window. Down the left hand side of the New Data Window the contained numerical data represents the number of respondents from one to one hundred.

The ordinal data, i.e. questions one to twelve are coded systematically. For example, in question one, Male is coded 1 and Female is coded 2. In question two the age groups are coded as follows, 16 to 19 is coded 1, 20 to 23 is coded 2, 24 to 27 is coded 3, 28 to 31 is coded 4, 32 to 35 is coded 5, 36 to 39 is coded 6, and over 40 is coded 7. This system of coding is continued for the remaining ordinal questions.

The interval data questions from thirteen to one hundred and twenty eight are coded so that the summated rated scale of measurement is given the number 1 at the left of the scale or the strongly disagree point to 5 at the strongly agree point.

To encode the open ended questions all of the data was correlated into rank categories, each category receiving a number starting with the number one representing category one and so on.

I have used descriptive and frequency tests on all questions.

SPSS is capable of performing the following statistical tests on the encoded data. Measurement of association which includes, Lambda, Gamma, Pearson’s r Product-Moment Coefficient, Regression.

Tests of significance including Chi-Square (used on a selected number of questions), t tests, analysis of Variance.

Bivariate and Multivariate analysis can be performed.

Pilot Survey

The questionnaire reliability was verified on a small random sample of twenty immediate work colleagues of Air Support Company Signals on Monday 16th September 1996. The purpose of conducting the pilot survey is outlined hereunder.

      1. Observe the level of respondent interest to the questions.
      2. Ensure that all members of the pilot sample understood the questions posed thus leaving no room for ambiguity.
      3. Test the question sequence.
      4. Establish continuity and flow between questions and division of categories.
      5. Consider survey question construction, length and timing (40 mins).

After suitable structure modification, the questionnaire instrument was distributed through Unit Orderly Rooms, on Monday 30th September 1996.

Response Rate of Return

Most completed questionnaires were returned during October 1996. A total of eighty-eight was received during this time.

During the month of November 1996, a total of twelve completed questionnaires were returned.

Several reasons precipitated the poor response rate which is outlined.

      1. Lack of interest in the organisation.
      2. Poor morale in the Air Corps.
      3. The perception of job insecurity.
      4. Organisational fear brought on by rumour and mistrust.
      5. Internalisation of organisational oppression.
      6. External consultants currently examining the role for the Air Corps into the 21st century.
      7. The concern regarding possible contracting, subcontracting of work out of the organisation.
      8. Management conducting a work activity survey among the workforce to be used by the external management consultancy firm. This survey took place two weeks before the release of my questionnaire.
      9. The fear among respondents that analysis emanating from my research would help the external consultants and Air Corps senior management in their work.

To increase the response rate, I met with groups of targeted respondents, to alley fears, on six occasions, with a limited amount of success. All respondents were informed to ignore the closing date on the cover letter.

In order to prepare for the commencement of structured analysis of data, I self-imposed a termination date for the 31st of December 1996. Since then no more completed questionnaires have been returned.

Response Rate

The response rate was 20% of the targeted sample or 100 questionnaires were returned, which I consider to be a satisfactory return. Analysis based on the return will be statistically significant. With a return of 11.80% or almost 12% of the whole eligible population this gives an 88% chance that it truly represents the views of the population based at Baldonnel.

*****

1 Kane E.- Doing Your Own Research (1991). – P. 94 to P. 97.