Thesis
Evaluating readability as a factor in information security policies
- Creator
- Rights statement
- Awarding institution
- University of Strathclyde
- Date of award
- 2019
- Thesis identifier
- T15350
- Person Identifier (Local)
- 101470582
- Qualification Level
- Qualification Name
- Department, School or Faculty
- Abstract
- Policies should be treated as rules or principles that individuals can readily comprehend and follow as a pre-requisite to any organisational requirement to obey and enact regulations. This dissertation attempts to highlight one of the important factors to consider before issuing any policy that staff members are required to follow. Presently, there is no ready mechanism for estimating the likely efficacy of such policies across an organisation. One factor that has a plausible impact upon the comprehensibility of policies is their readability. Researchers have designed a number of software readability metrics that evaluate how difficult a passage is to comprehend; yet, little is known about the impact of readability on the interpretation of information security policies and whether analysis of readability may prove to be a useful insight. This thesis describes the first study to investigate the feasibility of applying readability metrics as an indicator of policy comprehensibility through a mixed methods approach, with the formulation and implementation of a seven phase sequential exploratory fully mixed methods design. Each one was established in light of the outcomes from the previous phase. The methodological approach of this research study is one of the distinguishing characteristics reported in the thesis, which was as follows: * eight policies were selected (from a combination of academia and industry sector institutes); * specialists were requested their insights on key policy elements; * focus group interviews were conducted; * comprehension tests were developed (Cloze tests); * a pilot study of comprehension tests was organised (preceded by a small-scale test); * a main study of comprehension tests was performed with 600 participants and reduce that for validation to 396; * a comparison was made of comprehension results against readability metrics. The results reveal that the traditional readability metrics are ineffective in predicting human estimation. Nevertheless, readability, as measured using a bespoke readability metric, may yield useful insight upon the likely difficulty that end-users may face in comprehending a written text. Thereby, our study aims to provide an effective approach to enhancing the comprehensibility of information security policies and afford a facility for future research in this area. The research contributes to our understanding of readability in general and offering an optimal technique to measure the readability in particular. We recommend immediate corrective actions to enhance the ease of comprehension for information security policies. In part, this may reduce instances where users avoid fully reading the information security policies, and may also increase the likelihood of user compliance. We suggest that the application of appropriately selected readability assessment may assist policy makers to test their draft policies for ease of comprehension before policy release. Indeed, there may be grounds for a readability compliance test that future information security policies must satisfy.
- Advisor / supervisor
- Weir, George
- Resource Type
- Note
- This thesis was previously held under moratorium from 26/11/19 to 26/11/21
- DOI
- Date Created
- 2019
- Former identifier
- 9912769093302996
Relations
Items
| Thumbnail | Title | Date Uploaded | Visibility | Actions |
|---|---|---|---|---|
|
|
PDF of thesis T15350 | 2021-07-02 | Public | Download |