Examining the direct and indirect effect that individual and contextual risk factors have in predicting substance abuse
- Authors: Fouten, Elron Steve
- Date: 2024-10-11
- Subjects: Structural equation modeling , Substance abuse South Africa , Ecological Systems Theory , Path analysis (Statistics) , Personality , Confirmatory factor analysis
- Language: English
- Type: Academic theses , Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/467006 , vital:76806 , DOI https://doi.org/10.21504/10962/467006
- Description: Substance abuse levels in South Africa have continued to rise, with the age of first experimentation with drugs reported as being 10 years. Several studies have shown that substance abuse has an adverse impact on users, their families and their communities, resulting in a number of social, psychological and economic struggles. Moreover, substance abuse places an increased psychological, social and financial burden on the individual and the family, as it has been related to the destabilisation of the nuclear and extended family units, permeating every area of life and affecting the very social fabric of society. This study therefore endeavoured to determine which of the individual and contextual risk factors measured by the South African Substance Use Contextual Risk Instrument (SASUCRI) best predicts substance abuse in a sample of self-identified substance abusers. Additionally, the study sought to determine the magnitude, strength and direction of the interaction of the individual and contextual risk factors in predicting substance abuse. The specific aims of the study were: 1) to build a theoretical model that best approximates the identified and measured individual and contextual factors associated with substance abuse, 2) to use structural equation modelling (SEM) to test the adequacy of the model’s fit to the data that have been collected, 3) to use SEM to statistically determine which of the perceived individual or contextual factors best predict substance abuse, and 4) to use SEM to statistically determine which combination of perceived individual and contextual factors best predict substance abuse. To achieve these aims, the study applied SEM to data collected from individuals seeking either inpatient or outpatient treatment for substance abuse in the Eastern Cape and Western Cape Provinces of South Africa, respectively. The purpose of SEM is to a priori specify a theoretical causal model consisting of a set of predicted covariances between variables, and then test whether it is plausible when compared to the observed data. The appropriateness of performing SEM exists in its ability and suitability to examine the nature and magnitude of postulated dependence relationships, while simultaneously assessing the direct and indirect relations of the variables. The theoretical causal model that was tested contained latent and manifest variables that were identified as risk factors for substance abuse. The causal model was informed by Bronfenbrenner’s (1977a, 1986) ecological systems theory (EST) and was specified to approximate the different systems of the theory. Data for the study were collected using the relatively newly developed and validated SASUCRI. The results of the confirmatory factor analysis (CFA) showed that the SASUCRI was a reliable and valid instrument to use with this population, and that the results of the structural model can be interpreted with confidence. The fit statistics, for the normal theory and related bootstrap latent path SEM, all indicate that the model is an excellent fit to the data CMIN/DF = 2.82, NFI = .946, CFI = .97, RMSEA = .04., thus achieving the study’s objective. The normal theory analysis of the paths in the diagram identified the microsystem, “Access/proximity to drugs” (APTD) and “Concerns for future” (CFF) as significant predictors of substance abuse, whereas the bootstrap analysis of the paths identified the macrosystem, APTD and CFF as significant predictors of substance abuse. This study identified 10 risk profiles that interact in the prediction of substance abuse; these are: individuals with low self-efficacy (SE) who perceive their family as lacking or having low intra-family communication, who lack positive peer support, who perceive criminal behaviours as normal, who feel a strong need to fit in, who are less religious, who have easy access to drugs, who lack a sense of community traditions and belonging, who see public displays of substance use (SU) as normal, and who lack concern for the future, are at 59% greater risk of substance abuse. Though micro level factors (low SE, lack of or low intra-family communication, and a lack of positive peer support) and meso-level factors (normalisation of criminal activity, a strong need to fit in, and being less religious) does not significantly predict abuse, they significantly interact with external factors in predicting substance abuse. Thus, it is only when they interact with factors located in the other systems (exosystem, macrosystem and chronosystem) that they become significant predictors of substance abuse. Conversely, exosystem (easy access to drugs), macrosystem (lack of a sense of community traditions and belonging, commonplace public displays of SU), and chronosystem factors (lack of concern for the future), independently are significant predictors of substance abuse. These findings carry significant public health implications by challenging the prevailing focus on individual-based interventions. Recognising that the drivers of substance abuse extend beyond individual factors, this study therefore advocates for interventions that address the broader systems contributing to the issue. This has the potential to inform more effective and holistic public health strategies. Finally, this study emphasises the need for comprehensive strategies that span different systems, acknowledging the role of family, community and societal factors. This call for complex interventions aligns with the ecological systems perspective, advocating for a paradigm shift in how we address SU by considering the broader environmental influences that contribute to the problem. In summary, this study not only contributes to the academic understanding of SU, but also has practical implications for public health policies and interventions. , Thesis (PhD) -- Faculty of Humanities, Psychology, 2024
- Full Text:
- Authors: Fouten, Elron Steve
- Date: 2024-10-11
- Subjects: Structural equation modeling , Substance abuse South Africa , Ecological Systems Theory , Path analysis (Statistics) , Personality , Confirmatory factor analysis
- Language: English
- Type: Academic theses , Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/467006 , vital:76806 , DOI https://doi.org/10.21504/10962/467006
- Description: Substance abuse levels in South Africa have continued to rise, with the age of first experimentation with drugs reported as being 10 years. Several studies have shown that substance abuse has an adverse impact on users, their families and their communities, resulting in a number of social, psychological and economic struggles. Moreover, substance abuse places an increased psychological, social and financial burden on the individual and the family, as it has been related to the destabilisation of the nuclear and extended family units, permeating every area of life and affecting the very social fabric of society. This study therefore endeavoured to determine which of the individual and contextual risk factors measured by the South African Substance Use Contextual Risk Instrument (SASUCRI) best predicts substance abuse in a sample of self-identified substance abusers. Additionally, the study sought to determine the magnitude, strength and direction of the interaction of the individual and contextual risk factors in predicting substance abuse. The specific aims of the study were: 1) to build a theoretical model that best approximates the identified and measured individual and contextual factors associated with substance abuse, 2) to use structural equation modelling (SEM) to test the adequacy of the model’s fit to the data that have been collected, 3) to use SEM to statistically determine which of the perceived individual or contextual factors best predict substance abuse, and 4) to use SEM to statistically determine which combination of perceived individual and contextual factors best predict substance abuse. To achieve these aims, the study applied SEM to data collected from individuals seeking either inpatient or outpatient treatment for substance abuse in the Eastern Cape and Western Cape Provinces of South Africa, respectively. The purpose of SEM is to a priori specify a theoretical causal model consisting of a set of predicted covariances between variables, and then test whether it is plausible when compared to the observed data. The appropriateness of performing SEM exists in its ability and suitability to examine the nature and magnitude of postulated dependence relationships, while simultaneously assessing the direct and indirect relations of the variables. The theoretical causal model that was tested contained latent and manifest variables that were identified as risk factors for substance abuse. The causal model was informed by Bronfenbrenner’s (1977a, 1986) ecological systems theory (EST) and was specified to approximate the different systems of the theory. Data for the study were collected using the relatively newly developed and validated SASUCRI. The results of the confirmatory factor analysis (CFA) showed that the SASUCRI was a reliable and valid instrument to use with this population, and that the results of the structural model can be interpreted with confidence. The fit statistics, for the normal theory and related bootstrap latent path SEM, all indicate that the model is an excellent fit to the data CMIN/DF = 2.82, NFI = .946, CFI = .97, RMSEA = .04., thus achieving the study’s objective. The normal theory analysis of the paths in the diagram identified the microsystem, “Access/proximity to drugs” (APTD) and “Concerns for future” (CFF) as significant predictors of substance abuse, whereas the bootstrap analysis of the paths identified the macrosystem, APTD and CFF as significant predictors of substance abuse. This study identified 10 risk profiles that interact in the prediction of substance abuse; these are: individuals with low self-efficacy (SE) who perceive their family as lacking or having low intra-family communication, who lack positive peer support, who perceive criminal behaviours as normal, who feel a strong need to fit in, who are less religious, who have easy access to drugs, who lack a sense of community traditions and belonging, who see public displays of substance use (SU) as normal, and who lack concern for the future, are at 59% greater risk of substance abuse. Though micro level factors (low SE, lack of or low intra-family communication, and a lack of positive peer support) and meso-level factors (normalisation of criminal activity, a strong need to fit in, and being less religious) does not significantly predict abuse, they significantly interact with external factors in predicting substance abuse. Thus, it is only when they interact with factors located in the other systems (exosystem, macrosystem and chronosystem) that they become significant predictors of substance abuse. Conversely, exosystem (easy access to drugs), macrosystem (lack of a sense of community traditions and belonging, commonplace public displays of SU), and chronosystem factors (lack of concern for the future), independently are significant predictors of substance abuse. These findings carry significant public health implications by challenging the prevailing focus on individual-based interventions. Recognising that the drivers of substance abuse extend beyond individual factors, this study therefore advocates for interventions that address the broader systems contributing to the issue. This has the potential to inform more effective and holistic public health strategies. Finally, this study emphasises the need for comprehensive strategies that span different systems, acknowledging the role of family, community and societal factors. This call for complex interventions aligns with the ecological systems perspective, advocating for a paradigm shift in how we address SU by considering the broader environmental influences that contribute to the problem. In summary, this study not only contributes to the academic understanding of SU, but also has practical implications for public health policies and interventions. , Thesis (PhD) -- Faculty of Humanities, Psychology, 2024
- Full Text:
A personality-based behavioural model: Susceptibility to phishing on social networking sites
- Authors: Frauenstein, Edwin Donald
- Date: 2021-10-29
- Subjects: Phishing , Social networks , Personality , Self-presentation in mass media , Internet fraud , Internet users Habits and behavior , Big Five model , Human information processing , Heuristic-Systematic Model (HSM)
- Language: English
- Type: Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/190306 , vital:44982 , 10.21504/10962/190306
- Description: The worldwide popularity of social networking sites (SNSs) and the technical features they offer users have created many opportunities for malicious individuals to exploit the behavioral tendencies of their users via social engineering tactics. The self-representation and social interactions on SNSs encourage users to reveal their personalities in a way which characterises their behaviour. Frequent engagement on SNSs may also reinforce the performance of certain activities, such as sharing and clicking on links, at a “habitual” level on these sites. Subsequently, this may also influence users to overlook phishing posts and messages on SNSs and thus not apply sufficient cognitive effort in their decision-making. As users do not expect phishing threats on these sites, they may become accustomed to behaving in this manner which may consequently put them at risk of such attacks. Using an online survey, primary data was collected from 215 final-year undergraduate students. Employing structural equation modelling techniques, the associations between the Big Five personality traits, habits and information processing were examined with the aim to identify users susceptible to phishing on SNSs. Moreover, other behavioural factors such as social norms, computer self-efficacy and perceived risk were examined in terms of their influence on phishing susceptibility. The results of the analysis revealed the following key findings: 1) users with the personality traits of extraversion, agreeableness and neuroticism are more likely to perform habitual behaviour, while conscientious users are least likely; 2) users who perform certain behaviours out of habit are directly susceptible to phishing attacks; 3) users who behave out of habit are likely to apply a heuristic mode of processing and are therefore more susceptible to phishing attacks on SNSs than those who apply systematic processing; 4) users with higher computer self-efficacy are less susceptible to phishing; and 5) users who are influenced by social norms are at greater risk of phishing. This study makes a contribution to scholarship and to practice, as it is the first empirical study to investigate, in one comprehensive model, the relationship between personality traits, habit and their effect on information processing which may influence susceptibility to phishing on SNSs. The findings of this study may assist organisations in the customisation of an individual anti-phishing training programme to target specific dispositional factors in vulnerable users. By using a similar instrument to the one used in this study, pre-assessments could determine and classify certain risk profiles that make users vulnerable to phishing attacks. , Thesis (PhD) -- Faculty of Commerce, Information Systems, 2021
- Full Text:
- Authors: Frauenstein, Edwin Donald
- Date: 2021-10-29
- Subjects: Phishing , Social networks , Personality , Self-presentation in mass media , Internet fraud , Internet users Habits and behavior , Big Five model , Human information processing , Heuristic-Systematic Model (HSM)
- Language: English
- Type: Doctoral theses , text
- Identifier: http://hdl.handle.net/10962/190306 , vital:44982 , 10.21504/10962/190306
- Description: The worldwide popularity of social networking sites (SNSs) and the technical features they offer users have created many opportunities for malicious individuals to exploit the behavioral tendencies of their users via social engineering tactics. The self-representation and social interactions on SNSs encourage users to reveal their personalities in a way which characterises their behaviour. Frequent engagement on SNSs may also reinforce the performance of certain activities, such as sharing and clicking on links, at a “habitual” level on these sites. Subsequently, this may also influence users to overlook phishing posts and messages on SNSs and thus not apply sufficient cognitive effort in their decision-making. As users do not expect phishing threats on these sites, they may become accustomed to behaving in this manner which may consequently put them at risk of such attacks. Using an online survey, primary data was collected from 215 final-year undergraduate students. Employing structural equation modelling techniques, the associations between the Big Five personality traits, habits and information processing were examined with the aim to identify users susceptible to phishing on SNSs. Moreover, other behavioural factors such as social norms, computer self-efficacy and perceived risk were examined in terms of their influence on phishing susceptibility. The results of the analysis revealed the following key findings: 1) users with the personality traits of extraversion, agreeableness and neuroticism are more likely to perform habitual behaviour, while conscientious users are least likely; 2) users who perform certain behaviours out of habit are directly susceptible to phishing attacks; 3) users who behave out of habit are likely to apply a heuristic mode of processing and are therefore more susceptible to phishing attacks on SNSs than those who apply systematic processing; 4) users with higher computer self-efficacy are less susceptible to phishing; and 5) users who are influenced by social norms are at greater risk of phishing. This study makes a contribution to scholarship and to practice, as it is the first empirical study to investigate, in one comprehensive model, the relationship between personality traits, habit and their effect on information processing which may influence susceptibility to phishing on SNSs. The findings of this study may assist organisations in the customisation of an individual anti-phishing training programme to target specific dispositional factors in vulnerable users. By using a similar instrument to the one used in this study, pre-assessments could determine and classify certain risk profiles that make users vulnerable to phishing attacks. , Thesis (PhD) -- Faculty of Commerce, Information Systems, 2021
- Full Text:
A personality-based surveillance model for Facebook apps
- Authors: Van der Schyff, Karl Izak
- Date: 2020
- Subjects: Facebook (Electronic resource) , Electronic surveillance -- Psychological aspects , Online social networks -- Psychological aspects , Social media -- Psychological aspects , Personality
- Language: English
- Type: text , Thesis , Doctoral , PhD
- Identifier: http://hdl.handle.net/10962/145534 , vital:38447
- Description: The surveillance of data through the use of Facebook Apps is an ongoing and persistent problem that impacts millions of users. Nonetheless, limited research has been conducted investigating to what extent a Facebook user’s personality influences their awareness of such surveillance practices. Thus, to understand this situation better, the current study inductively developed four propositions from secondary data sources as part of a detailed content analysis. Spanning three search and analysis phases the content analysis led to the development of the research model. Guided by the propositions and research questions, a questionnaire was developed based on the relevant constructs prescribed by the Theory of Planned Behaviour. This questionnaire was used, and a total of 651 responses were collected from Facebook users over the age of 18 years old and residing in the United States of America. Primary data took place at both a univariate and multivariate level with a specific focus on the development of a structural model. Interpretation of the structural model revealed that out of all the Big Five personality traits, Conscientiousness exhibited the strongest relationship with information security awareness followed by Openness to Experience and Neuroticism, respectively. The results further indicated that the model constructs based on attitude, social norms and awareness significantly influenced the intended use of Facebook Apps. The study also contributes by indicating which personality traits are most vulnerable to Facebook App surveillance. For example, it was found that individuals high in Conscientiousness are the least vulnerable with individuals high in Extraversion being the most vulnerable. Since the results indicate that not all the personality traits are significantly related to the model constructs, additional factors may contribute to App surveillance in this context. Concerning this, factors such as user apathy, information privacy, privacy concerns, control and Facebook dependency are discussed as a means to argue why this might be the case.
- Full Text:
- Authors: Van der Schyff, Karl Izak
- Date: 2020
- Subjects: Facebook (Electronic resource) , Electronic surveillance -- Psychological aspects , Online social networks -- Psychological aspects , Social media -- Psychological aspects , Personality
- Language: English
- Type: text , Thesis , Doctoral , PhD
- Identifier: http://hdl.handle.net/10962/145534 , vital:38447
- Description: The surveillance of data through the use of Facebook Apps is an ongoing and persistent problem that impacts millions of users. Nonetheless, limited research has been conducted investigating to what extent a Facebook user’s personality influences their awareness of such surveillance practices. Thus, to understand this situation better, the current study inductively developed four propositions from secondary data sources as part of a detailed content analysis. Spanning three search and analysis phases the content analysis led to the development of the research model. Guided by the propositions and research questions, a questionnaire was developed based on the relevant constructs prescribed by the Theory of Planned Behaviour. This questionnaire was used, and a total of 651 responses were collected from Facebook users over the age of 18 years old and residing in the United States of America. Primary data took place at both a univariate and multivariate level with a specific focus on the development of a structural model. Interpretation of the structural model revealed that out of all the Big Five personality traits, Conscientiousness exhibited the strongest relationship with information security awareness followed by Openness to Experience and Neuroticism, respectively. The results further indicated that the model constructs based on attitude, social norms and awareness significantly influenced the intended use of Facebook Apps. The study also contributes by indicating which personality traits are most vulnerable to Facebook App surveillance. For example, it was found that individuals high in Conscientiousness are the least vulnerable with individuals high in Extraversion being the most vulnerable. Since the results indicate that not all the personality traits are significantly related to the model constructs, additional factors may contribute to App surveillance in this context. Concerning this, factors such as user apathy, information privacy, privacy concerns, control and Facebook dependency are discussed as a means to argue why this might be the case.
- Full Text:
The effect of perceptual skill of RFT scores : a cross-cultural study
- Authors: Marais, W F
- Date: 1977
- Subjects: Witkin, Herman A , Personality , Personality and culture -- Cross-cultural studies , Cross-cultural studies , Perception
- Language: English
- Type: Thesis , Masters , MSocSc
- Identifier: vital:3285 , http://hdl.handle.net/10962/d1012838
- Description: The increasing need to select individuals for tasks suited to their personality make-up, has added to the challenge psychology faces of developing tests which can be applied to subjects from different cultures and environments. Many attempts at such designs litter the history of industrial and cross- cultural personality research. Among those to have survived years of reassessment, is the cognitive style approach developed by Herman Witkin. Intro., p. 1.
- Full Text:
- Authors: Marais, W F
- Date: 1977
- Subjects: Witkin, Herman A , Personality , Personality and culture -- Cross-cultural studies , Cross-cultural studies , Perception
- Language: English
- Type: Thesis , Masters , MSocSc
- Identifier: vital:3285 , http://hdl.handle.net/10962/d1012838
- Description: The increasing need to select individuals for tasks suited to their personality make-up, has added to the challenge psychology faces of developing tests which can be applied to subjects from different cultures and environments. Many attempts at such designs litter the history of industrial and cross- cultural personality research. Among those to have survived years of reassessment, is the cognitive style approach developed by Herman Witkin. Intro., p. 1.
- Full Text:
- «
- ‹
- 1
- ›
- »