In recent years, a lack of reproducibility in science has come to light. One of the causes of this lack of reproducibility is the use of questionable research practices (QRPs). P-hacking is one of...Show moreIn recent years, a lack of reproducibility in science has come to light. One of the causes of this lack of reproducibility is the use of questionable research practices (QRPs). P-hacking is one of these QRPs and consists of making impromptu data analytic choices, resulting in a biased significant p-value. The practice of blind analysis could be a viable option to tackle p-hacking. This thesis introduces a new categorization of blind analysis-methods: label blinding, noise blinding and coefficient blinding. Each of these categories has its own implications when it comes to the data analysis. Label blinding allows a researcher to check assumptions with the blinded data, but does not hide the actual p-values. Noise blinding prevents a researcher from checking assumptions with the blinded data, but hides the actual p¬-values. Coefficient blinding allows a researcher to check assumptions and hides the actual p-values, but requires a researcher to specify the model they are going to fit beforehand. Three case studies with simulated data illustrate what the plots and tests used for assumption checks look like for a method from each blind analysis category, as well as what the test statistics might look like. When using a label blinding method while conducting a two-way ANOVA, we are able to check the assumptions of a two-way ANOVA, but our results are not hidden, leaving an opening for p-hacking. When using a noise blinding method while conducting a two sample t-test, we are unable to check the assumptions of a two sample t-test, but our test results are hidden. When using a coefficient blinding method while conducting a linear regression analysis, we are able to check the assumptions of a linear regression and our test results are hidden. Not every blind analysis method is applicable to every dataset or data analysis. More research is needed to explore what blind analysis methods are applicable to which tests. Additionally, guidelines for the practical implementation of blind analysis should be specified.Show less
Human beings learn about themselves through social information, meaning that their self-esteem is dependent on self-referencing feedback provided by their social environment. While self-esteem has...Show moreHuman beings learn about themselves through social information, meaning that their self-esteem is dependent on self-referencing feedback provided by their social environment. While self-esteem has been comprehensively studied, the concept of self-esteem contingency (i.e., how strongly individuals update their self-esteem in response to social ego-referencing feedback) has yet to be investigated systematically. More specifically, while we have a limited understanding of this construct in adults, its expression in adolescents remains a mystery. However, developmental studies indicate that self-esteem dynamics in adults and adolescents vary, emphasising the importance of investigating this construct in all age groups. Therefore, this study examined whether adults and adolescents show differences in self-esteem contingency by investigating whether self-esteem, age, and sex influence this construct. One hundred and six adults and 64 adolescents were tasked with performing a speech in front of three alleged judges. Participants subsequently evaluated themselves and received pre-generated positive or negative feedback about their performance. Computational modelling was used to assess trial-to-trial changes in self-esteem contingency. Results revealed that adults processed positive and negative feedback very distinctly, while only a small inclination towards such two-dimensional strategies was observed in adolescents. Moreover, the data illustrated a positivity bias in adults with high self-esteem, whereas a negativity bias was found in adolescents with lower self-esteem. Furthermore, adolescents and adults did not differ in self-esteem contingency and proved more contingent on positive than negative feedback. Lastly, we found no impact of the participants’ sex on their self-esteem contingency in either age group. Overall, these findings indicate that adults have developed a self-protective strategy that allows them to incorporate positive information more strongly into their self-view, whereas this skill seems less sophisticated in adolescents.Show less
Almost 90 % of the archaeological sites in Barbados are located within one kilometre from the coast, implying some sort of preference towards it. Some scholars suggest this might be because the pre...Show moreAlmost 90 % of the archaeological sites in Barbados are located within one kilometre from the coast, implying some sort of preference towards it. Some scholars suggest this might be because the pre-colonial inhabitants of the island actually favoured those places, while concerns about a possible bias have been raised by others. Reasons for the latter become more obvious when looking at the nature of exploration; the coast has been extensively investigated, while inland research is often neglected. Due to this limited exploration, certain concepts and ideas about pre-colonial lifeways could become one-sided and biased. This is visible through a distinct pattern in literature, restricting pre-colonial ways of life to the shore and even excluding the work of other scholars, furthering the incompleteness of the research. Fortunately, various efforts have been made recently to address these issues, but it remains hard to grasp the scope of and reasoning behind this bias. Therefore, this research aims to study this bias by summarising acknowledged and unacknowledged biases from the archaeological literature on the island. Additionally, an overview of possible inland sites that were previously not mapped is created to further support the idea of a bias towards the coast. These include the so-called ‘Indian’ places, probably named after their former presence, registrations of artefacts at, for example, historical excavations, and also ecological factors that played a vital role in pre-colonial lifestyles. The results of this research show a clear relationship between potential inland sites and sources of freshwater, especially big rivers. There is a greater density of possible inland sites in the northern half of the island compared to the south, but this may be the result of the bias. Although actual fieldwork is needed to confirm the pre-colonial nature of these places, it clearly demonstrates a preference towards coastal archaeology by scholars rather than the pre-colonial inhabitants. Vastly limited by the lack of fieldwork and elaborate reporting, it remains hard to conclude with certainty how the story of those inhabitants would exactly change. However, it can be concluded that the general narrative focussed on the idea that pre-colonial life was mainly centred on the coast is probably more diverse.Show less
This study had the objective to see the effect between the gender of a police officer and their likelihood to report on domestic violence. In order to answer the research question, “To what extent...Show moreThis study had the objective to see the effect between the gender of a police officer and their likelihood to report on domestic violence. In order to answer the research question, “To what extent does gender affect a police officer’s likelihood to file a report on domestic violence?”, an experimental survey was conducted, which was focussed on the likelihood to report as well as the credibility of the victim. It was expected that female police officers would be more likely to report alleged domestic violence if the victim was a woman and male police officers were expected to be more likely to report if the victim was male, due to their match of gender, which was based on the representative bureaucracy theory. For conclusion, it can be stated that this hypothesis (H1) is not true, as both genders tend to be more likely to report if the victim is a woman. Moreover, it was expected that female victims would be more credible to both male and female officers. When analysing the means of the outcomes, the second hypothesis (H2) seemed true. However, reliability has been a concern as the data is not statistically significant, meaning that for both statements, the null hypothesis cannot be rejected. This has occurred because of a Type II error, due to insufficient sampling. The ration men-women has not been executed well, resulting in to less male participants compared to female participants.Show less
Longitudinal data are often collected in different research areas such as medicine, biology, education, and psychology. We can build a transitional model using longitudinal binary data, which aims...Show moreLongitudinal data are often collected in different research areas such as medicine, biology, education, and psychology. We can build a transitional model using longitudinal binary data, which aims to model the probability of transition between the response categories. In this type of data is common to find missing values due to dropouts, costs, and organizational problems. The missing-indicator model is often used as a method to handle missing values in this type of data. This method consists in creating a new category for the missing values. Therefore, the binary logistic model changes to a baseline-category logit model. This study aims to evaluate the bias of the estimated coefficients when the missing-indicator method is used in the response of a binary transitional model. Based on an empirical example, a Monte Carlo simulation with three factors is carried out: (1) type of missingness, (2) sample size, and (3) proportion of missing data. The coefficients bias from the baseline-category logit model is evaluated using boxplots and a three-way MANOVA analysis. The results suggest that sample size, the proportion of missing data, type of missingness and the interaction between sample size and proportion affect the bias of the estimated coefficients; nonetheless, the effect size is small. When each dependent variable is analysed separately using ANOVA, the effects of the proportion of missing and the interaction between sample size and proportion were statistically significant for only one coefficient. However, the effect size is still small. Therefore, the conclusion is that the estimated coefficients' bias for all the missingness types is low.Show less
Smartphones have become an integral part of our daily lives. Along with this integration comes a concern for smartphone and internet addictions. In order to fully understand how these addictions...Show moreSmartphones have become an integral part of our daily lives. Along with this integration comes a concern for smartphone and internet addictions. In order to fully understand how these addictions might work, it is imperative that we develop more accurate measures of smartphone behaviours. Past studies have often included selfreported questionnaires that gathered data about the duration of smartphone use and other smartphone behaviours, but it is now known that self-reports are not entirely reliable. This study will investigate the difference between the self-reported duration of smartphone use and the actual duration provided by the users’ smartphones. Additionally, the study will explore the possibility of habitual checking behaviours as a predictor of recall error. Data collected from 122 participants were analysed via one-sample t-tests and multiple linear regressions. Results from the one-sample ttests support previous research, showing that participants were indeed inaccurate at recalling the time spent on their phones. Thus, smartphones are an important tool for providing objective data on smartphone behaviour. Contrary to former research, the level of smartphone usage did not make a difference on the amount of recall error, such that participants were inaccurate regardless of actual time spent on their smartphones. The multiple linear regressions found a relationship between one measure of time-based checking habits and recall error, but not between tap-based checking habits and recall error. These results indicate that unconscious, automatic smartphone habits may play a role in the inability to accurately recall smartphone behaviour.Show less
