Latent Class Models

A latent class methodology for conjoint analysis is proposed, which simultaneously estimates market segment membership and part-worth utilities for each derived market segment using mixtures of multivariate conditional normal distributions. An E-M algorithm to estimate the parameters of these mixtures is briefly discussed. Finally, an application of the methodology to a commercial study (pretest) examining the design of a remote automobile entry device is presented.

This paper investigates whether fathers who have been convicted of a violent offense transmit criminal and violent behavior more strongly than fathers who were convicted, but never for violence. First, a more traditional approach was taken where offending fathers were divided into two groups based on whether they had a violence conviction. Secondly, Latent Class Analysis (LCA) was performed to identify two classes of fathers, one of which was characterized as violent. Sons of fathers in this class had a higher risk of violent convictions compared with sons whose fathers were in the other class.

... eg, when a rumor has been spread) or remains anonymous (eg, as with cyberbullying). ...victimization would be associated with a pattern of multiple responses to victimization, because these ... The third aim of this study was to examine the associations between the response ...

W rozdziale opisane zostaną wybrane metody związane ze skończonymi modelami mieszanek (finite mixture models) ze szczególnym uwzględnieniem analizy klas ukrytych (latent class analysis, LCA), będącej jednym z przykładów takich modeli (Everitt i Hand, 1981; Wolfe, 1970). Ta grupa metod pozwala na pomiar zmiennych ukrytych o charakterze dyskretnym. Podstawowym ich założeniem omawianych jest istnienie heterogenicznej populacji, z której zostali wylosowani respondenci. Można ją podzielić na pewną liczbę homogenicznych subpopulacji, wykorzystując do tego model statystyczny. Polega to na wyodrębnianiu z danych obserwowalnych zmiennej ukrytej, definiującej przydział do poszczególnych skupień, charakteryzujących się odrębnymi rozkładami. Podstawowym zastosowaniem tych metod jest analiza skupień, z tego powodu bywają nazywane analizą skupień opartą na modelu (...)

Testing for measurement invariance can be done within the context of multigroup latent class analysis. Latent class analysis can model any type of discrete level data, which makes it an obvious choice when nominal indicators are used or when a researcher's aim is at classifying respondents in latent classes. The multigroup latent class (LC) model can be specified in three different ways, i.e. by adopting a probabilistic, a log-linear or a logistic parameterization. We define and compare these different forms of parameterization. The starting point is the standard LC model in which indicators and latent variables are defined at the nominal level. Additionally, we focus on LC models with ordinal indicators as well as LC factor models with ordinal indicators. Testing for measurement invariance involves estimating LC models with different degrees of homogeneity. We explain the procedure for investigating measurement invariance at both the scale as well as the item level. We illustrate the approach with two examples. The first example is a multigroup LC analysis with nominal indicators; the second a multigroup LC factor analysis with ordinal indicators.

The tetrachoric correlation coefficient measures the relation of a 2x2 contingency table, assuming that each categorical variable has an underlying continuous normal distribution. The iteration procedure utilizes tetrachoric series expansion and Gaussian quadrature, depending upon whether the absolute value of the estimate exceeds .95. This article describes a SAS/IML program that calculates the tetrachoric correlation coefficient and its asymptotic standard error. It also performs a t-test for the hypothesis Ho: ρ=0.

Rating scales are popular for measuring attitudes, but response style, a source of measurement error associated with this type of question, can result in measurement bias in important attitudinal measures. Although numerous research efforts have been devoted to this topic, there are still some overlooked areas. This dissertation intends to fill three gaps in the literature on response style. Chapter 1 investigates the effects of face-to-face and Web survey on acquiescent response style (ARS) and extreme response style (ERS) using the 2012 American National Election Studies (ANES) data. Using the latent class analysis approach, I find that: 1) both ARS and ERS exist in both face-to-face and Web survey; 2) face-to-face respondents demonstrate more ARS and ERS than Web respondents; 3) the effect of mode on ERS is larger for black respondents than for white and Hispanic respondents. Chapter 2 compares ERS with respect to the format of response scale, specifically agree-disagree (A/D) and item specific (IS) scales. This study analyzes a between- and within-subject experiment embedded in the 2012 ANES. Using latent class factor analysis, I reached the following three major findings: 1) ERS exists in both A/D and IS scale formats; 2) ERS shows a slightly different pattern between the two scale formats; 3) when analyzing ERS within subjects across two waves, there is only a single ERS latent class variable for both scale formats, after controlling for the correlation within respondents. Chapter 3 utilizes the 2003 Detroit Arab American Study to examine the impact of acculturation of Arab Americans on ERS. The results indicate that less acculturated respondents are more prone to ERS than more acculturated respondents, and this is especially true for the
5-point rating response scales as compared to 3-point rating scales. This phenomenon can be explained by the fact that less acculturated respondents identify more strongly with honor-based collectivist cultures that value decisive and assertive answers since this is a way of showing one’s unambiguous attitude and standing, an important quality in such a culture. The language of the interview primes the relevant cultural norms and therefore mediates the relationship between acculturation and ERS.

The present study examined the extent to which scores on the Dispositional Flow Scale-2 (DFS-2) could differentiate individuals who frequently experience flow characteristics in physical activity from those who do not. A total of 993 participants completed the Japanese version of the DFS-2. Latent class factor analysis (LCFA), which combines the strengths of both latent class analysis and factor analysis, was conducted on the DFS-2 responses. Six classes were identified through a series of LCFAs and the patterns of the item-average scores for the nine flow attributes were found to be parallel among these classes. The top two and bottom two classes (19.3% and 13.4% of the whole sample) were considered the groups who experience flow characteristics frequently and seldom, respectively. These results indicated that individuals who often experience flow attributes in physical activity could be differentiated from those who do not based on their DFS-2 scores. Note. The article published i...

Public policies should reflect and accommodate as much as possible citizens’ preferences and values. However, it is dicult to know the correct value citizens place on public goods as they are not generally exchanged in the marketplace. For this reason,
non market valuation is increasingly considered to be an important tool for informing policy decisions. In this context, Contingent Valuation (CV) and Discrete Choice Experiments (DCE) are probably the most widely used tools to analyse Stated preferences and retrieve people’s Willingness to Pay (WTP) for improving a particular environmental amenity. Since its introduction as a technique of non-market valuation, the number of CV studies published in peer reviewed journals consistently increased for more than three decades. Then, since the mid to late 1990s, the CV technique has been largely substituted by the DCE method.

Having reviewed the literature, it was decided to focus my Phd on DCE method. Notwithstanding the importance of CV as a useful tool for preference elicitation, DCE proved more appealing, as recent literature have identified a number of exciting research
avenues from a methodological and econometric point of view.
Therefore, the PhD has been mainly focused on developing new tools for accommodating heterogeneity in tastes, variances and heuristics employing Latent Class (LC) analysis. As a result, I formalised a class of LC models that can accommodate heteroscedasticity and/or heterogeneity (depending on assumptions and parameterisation) within each class as well as dierent heuristics across classes. The three essays collected
in this thesis represent the main outcome of this research.

More specifically the first paper, on urban land use, introduces class heterogeneity within a LC model. This is obtained by specifying a discrete mixture of sets of continuous distributions. The model is applied to both simulated and real data in order to
demonstrate its flexibility and the advantages for policy appraisal.

The second paper introduces and formalises the idea of a heteroscedastic LC model using data from a recreational site choice study elicited through stated preference methods
to compare various model specifications. Results show that model fit, welfare estimates and choice predictions are sensitive to the manner in which both types of heterogeneity are accommodated for. This is done in WTP-space to directly compare estimates from continuous and discrete mixture representations as well as to demonstrate the importance of including the scale parameter even under this reparameterisation.

In the third and final paper, on species preservation, the well known problem of preference and variance instability due to learning and fatigue in DCE is tackled by applying a scale-adjusted latent class model to uncover both types of instability simultaneously and probabilistically across the sample. Findings highlight the advantages, in terms of model fit, interpretation and policy implications, that can be achieved when both types
of instability are addressed concurrently. Data collected to estimate the existence value of rare and endangered fish species in Ireland are utilised as an empirical case-study.

Finally conclusions and avenues for possible further research and future applications
of the presented models are drawn.

Il lavoro di tesi si colloca nell’ambito della ricerca sulla flessibilità del lavoro. Lo studio esplorativo della letteratura di riferimento ha messo in luce l’assenza di un metodo consolidato che consentisse un’analisi organica delle caratteristiche nel contesto italiano della flessibilità della prestazione di lavoro. Il lavoro è stato quindi condotto alla luce di questa sollecitazione. Da un punto di vista teorico si è scelto di riportare attenzione sulle prestazioni piuttosto che sulle tipologie contrattuali dei lavoratori, così da evidenziare la centralità che l’organizzazione del lavoro assume nella vita quotidiana dei lavoratori e delle loro famiglie. Con prestazione di lavoro si è inteso l'insieme di quelle caratteristiche che regolano le modalità temporali di svolgimento dell’attività lavorativa. Il focus dell’analisi è stato quindi posto sulle prestazioni flessibili di lavoro in relazione alle pratiche effettivamente osservate nel corso delle attività lavorative. Principale elemento di analisi è stata quindi l’esplorazione di dieci indicatori della flessibilità della prestazione, complessivamente intesa, individuabili nel dataset utilizzato. Per la realizzazione dell’intero complesso di analisi sono stati utilizzati i microdati Istat sulle Forze di lavoro 2009. I dieci indicatori utilizzati nella definizione del modello sono: organizzazione oraria contrattuale (part-time/full-time); il numero di ore settimanalmente lavorate (ricodificato in sei fasce); variazione (o meno) del monte orario dell’ultima settimana rispetto all’orario abituale; l’occorrenza (o meno) di lavoro straordinario; l’occorrenza (o meno) di lavoro notturno, serale, il sabato, la domenica, su turni, da casa.
Lo scopo dell’analisi è stato quindi di descrivere le forme flessibili della prestazione di lavoro in Italia attraverso una tipologia empirica. Per la definizione di questa tipologia si è fatto riferimento ad un approccio person-oriented dell’analisi della struttura latente. Pertanto la tipologia è stata realizzata con l’ausilio dell’analisi delle classi latenti (LCA) condotta per mezzo dell’applicativo R.
L’analisi ha consentito di identificare una tipologia della flessibilità della prestazione composta da quattro tipi di lavoratori flessibili: il tipo standard (57,6% sul totale dei lavoratori subordinati e parasubordinati), o tipo non flessibile, che identifica i soggetti con un probabilità elevata di non presentare nessuno dei caratteri definiti come flessibili; il tipo a bassa intensità lavorativa (12,8%) caratterizzato in modo flessibile per la probabilità di essere contrattualmente part-time ed avere , un monte orario ridotto; il tipo ad alta intensità lavorativa (15,5%) che, in modo opposto al tipo part-time, presenta un'alta probabilità di avere prestazioni caratterizzate da un elevato monte orario, da un'alta probabilità di lavoro straordinario e da picchi positivi di flessibilità oraria infrasettimanale; il tipo flessibile per organizzazione (14,0%), caratterizzato da alte probabilità di lavoro serale, notturno, di sabato, festivo e su turni.
La descrizione del mercato del lavoro italiano attraverso l’utilizzo della nuova tipologia è il risultato di natura descrittiva del lavoro svolto. Infine, principale risultato analitico, si è potuta verificare la sussistenza di una relazione statisticamente significativa tra la flessibilità contrattuale e la flessibilità della prestazione di lavoro. Le variabili di controllo inserite nel modello di regressione logistica multinomiale sono: il genere, la cittadinanza, il titolo di studio e l’età. I beta associati  sono 0,796 per il tipo a bassa intensità lavorativa, 0,511 per il tipo a alta intensità e 0,171 per il tipo flessibile per organizzazione.

This study investigates the psychometric properties, factor structure and validity of the revised Negative Acts  Auestionnaire-Revised (NAQ-R), an instrument designed to measure exposure to bullying in the workplace. By reanalyzing data based on a heterogeneous sample of 5288 UK employees, the results show that the 22-item instrument has a high internal stability,with three underlying factors: personal bullying, work-related bullying and physically intimidating
forms of bullying, although the instrument may also be used as a single factor measure. Criterion validity was explored by relating the scores on the NAQ-R to a single-item measure of perceived victimization from bullying, showing high correlations with both the total NAQ-R and scores on the three factors. Targets of bullying scored significantly higher on
all 22 items compared to non-targets. The NAQ-R correlated as expected with measures of mental health, psychosocial work environment and leadership, indicating a good construct
validity of the instrument. Furthermore, a latent class cluster (LCC) analysis showed that the instrument may be used to differentiate between groups of employees with different levels of exposure to bullying, ranging from infrequent exposure to incivility at work to severe victimization from bullying and harassment. The more commonly used operational criteria can also be used to detect targets of bullying. Hence, the NAQ-R is proposed as a standardized
and valid instrument for the measurement of workplace bullying.

This study characterises the ‘deal-proneness’ of consumers by analysis of the consumer-level characteristics of price sensitivity and brand loyalty. The study first develops a multinomial logistic (MNL) latent class model suitable for use with universal product code (UPC) point-of-sale (hypermarket) scanner data. The model is then used to assess the deal-proneness of consumers with respect to monetary promotions (price reductions) and non-monetary promotions (store flyers). The results show that almost 47% of consumers can be considered deal-prone, in that both kinds of sales promotions have a significant effect on their choice behaviour. The findings provide important insights for retail management in seeking to optimise the results obtained from promotional budgets.

This article investigates father and offspring criminal careers by employing the semi-parametric, group-based trajectories methodology. The findings demonstrate that children of sporadic and chronic offenders have significantly more convictions than children of non-offenders. However, contrary to expectations based on taxonomic and intergenerational theories, chronic offending fathers do not have more chronic offending children than sporadic fathers. The results demonstrate strong intergenerational transmission of criminal behaviour, but it is the fathers having a conviction rather than their conviction trajectory that is related to offspring convictions.

Workplace learning is a critical tool for both employees and organisations in their continuous adjustment to various global trends reshaping our workplaces, economies and societies. Workers learn most when engaging in new and challenging activities in which they are granted sufficient autonomy to solve issues themselves or in collaboration with their colleagues or mentors. Findings based on the data from the 2019 European Company Survey indicate that informal learning is the key aspect of workplaces in terms of company performance and employee wellbeing. Such results call for more attention from policy-makers and organisations’ management to various forms of informal learning at the workplace. This is particularly the case when considering the apparent lack of beneficial effects of formal and non-formal learning on employee wellbeing and company performance. These findings illustrate the overwhelming prominence of informal learning at the workplace compared to formal and non-formal learning forms. They also indicate the often untapped learning potential that exists at workplaces. However, analyses show that most companies are still not offering optimal learning environments and opportunities to their workers. Such a situation hurts both employees and companies but also presents an opportunity for action at organisational, national and international levels. Such an action would aim to create institutional conditions, legal frameworks, learning resources, and incentives for companies to start organising their workplaces in a way that will facilitate workplace learning, especially its various informal forms.