This research introduces a framework for explaining why managing family members create a shared understanding of caring and being cared for with their nonfamily employees, called ‘Social Reciprocity'. Applying an embedded case study design, using grounded theory method, this research adds to the understanding of the social mechanisms of internal stakeholder management in family firms, which have been largely ignored. The emergent framework proposes that the ability of family members to build trusting relationships with employees, based on cognitive and affective trusts contributes to the creation of social reciprocity. Affective trust, in particular, has a direct and moderating role in the creation of social reciprocity on emotional and instrumental level. The moderating role of affective trust is considered a valuable finding; since it emphasizes the role of trust as a meta theoretical concept. The developed framework suggests that affective trust initiates a social learning process that leads to a positive attribution bias, i.e. an automatic positive attribution of managing family members' actions as being based in benevolence and genuine care. Furthermore, social learning in the form of trust network reproduction emerges as contributory to the diffusion of social reciprocity to lower hierarchical levels. Analysis of data demonstrates that employees with little direct interactions with managing family members develop trusted weak ties to managing family members, with similar levels of cognitive and affective trust. This finding is particularly valuable as it challenges traditional network theory, which argues for frequent personal interaction to be necessary in creating trust ties. This thesis contributes to theoretical and methodological knowledge in several ways. First, it advances understanding of the social dynamics and mechanisms of internal stakeholder engagement in and SME family firm context. Second, the developed framework demonstrates the importance that concepts form multiple disciplines such as psychology, social psychology and sociology have on the development of theories in management research. Third, trust network reproduction and upper network stability emerged as causal mechanisms of potentially meta theoretical value that may have applicability on a wider range of topics in management research. Lastly, this thesis demonstrates the value of grounded theory in developing theory in management research.

The current study examined the role of a novel, adapted dialogic reading curriculum and its impact on preschoolers with autism and their interactions with their parents during shared book reading. The aim of this study was to determine the impact of the curriculum on the effects of child social reciprocity and parents’ feelings of competence and confidence when trained on implementation strategies. Pre- and post-test measures were conducted for four parent-child dyads to measure the impact of adapted dialogic reading on child social reciprocity and parents’ feelings of competence and confidence. Although the sample size was small, clear trends were seen suggesting adapted dialogic reading methods may result in greater increases in social reciprocity behaviors such as contingent responses to questions and joint attention during shared book reading. Positive trends also suggest that when parents are trained to implement adapted dialogic reading strategies, their feelings of competence and confidence are increased.

The topic of this diploma thesis is Interaction between parent and child with autistic spectrum disorder. The theoretical part is about mental development neurotypical preschool child and child with ASD, focused on social skills, speech and play. Emphasis is placed on reciprocity in parent-child socio-communication interaction. Next chapter is devoted to the specifics of raising a child with ASD. Empirical part contains qualitative research to investigate individual variability in social-communication behaviour in dyadic interaction parent-child with ASD. Videos were taken from which the socio-reciprocal skills, frequency and quality of interactions and initiation of contact on the part of the child and its response to parental initiation were analysed. Special attention was paid to making eye contact. The output of the thesis is practical recommendations for mothers, detailed qualitative analyses of observed interactions and a more detailed description of the socio-communicative behaviour of children with PAS. Keywords: interaction parent-child with ASD, social skills, social reciprocity, reciprocity in social communication, reciprocal-social behavior, play of preschool child, preschool child with ASD

Le Trouble du Spectre de l’Autisme (TSA) est un trouble neurodéveloppemental défini par deux caractéristiques principales : un comportement social atypique et des intérêts restreints et répétitifs incluant une rigidité comportementale. Le comportement social atypique est typiquement mesuré et quantifié par des déficits de Théorie de l’Esprit (ToM). Les outils classiques pour évaluer la ToM comportent plusieurs limites dont les biais liés à la subjectivité, la pauvre validité écologique et le manque d’échange réciproque. Ainsi, une tâche dyadique et intégrative pourrait pallier ces limites. En population adulte autiste, une telle tâche a été testée. Concrètement, le jeu du Penny-Hiding Game (PHG), où un joueur doit deviner dans laquelle de ses mains un adversaire a caché une pièce de monnaie, a été administré à un groupe de personnes autistes et un groupe de personnes non-autistes. La tâche comportait deux contextes : l’un social, où le participant croyait qu’il jouait contre un être humain en ligne, et l’un non-social, où le participant était informé qu’il jouait contre un ordinateur seulement. Cette étude a montré que les personnes autistes se montraient insensibles au contexte dans leur comportement modélisé par un phénotype computationnel comportant deux indices : la sophistication et la flexibilité. De plus, l’étude a identifié une stratégie de jeu qui semblait propre au groupe autiste. Le phénotype computationnel permettait aussi de classer les participants avec un taux de spécificité équivalent aux outils diagnostiques standards. La présente étude vise donc à étendre cette compréhension de la réciprocité sociale des personnes autistes à une population pédiatrique. L’enjeu des comorbidités en autisme constitue également un frein important aux avancées scientifiques. Ainsi, nous avons décidé d’inclure deux groupes dans notre échantillon qui présentent les comorbidités les plus prévalentes en autisme : le Trouble de l’Attention avec et sans Hyperactivité (TDAH) et les Troubles Anxieux (TA). À la suite d’une analyse de la performance avec ces trois groupes et une inspection visuelle des séquences de choix au cours de la tâche, un paramètre d’alternance/persévération et un groupe d’enfants neurotypiques (NT) ont été ajoutés. Ces deux ajouts font de ce mémoire une étude partiellement exploratoire. Le but était donc de valider la tâche informatique du PHG auprès de populations pédiatriques cliniques, d’identifier et d’explorer des différences de groupes au niveau des stratégies de jeu, de la performance et du phénotype computationnel et d’explorer le rôle potentiel des associations entre le phénotype computationnel, la performance et les symptômes associées aux trois conditions cliniques à l’étude : les symptômes autistiques, attentionnels/exécutifs et anxieux. Les analyses ont permis de révéler que le groupe NT avait une performance significativement supérieure aux groupes TSA, TDAH et TA et que la quantité et la sévérité des symptômes étaient négativement associées au score de performance en contexte social. La stratégie la plus utilisée à travers tous les groupes et indépendamment du contexte était une stratégie d’apprentissage par renforcement et aucune différence au niveau de la sophistication ou de la flexibilité n’a été détectée. Ces deux résultats constituent une non-réplication de ce qui a été observé chez les adultes. La persévérance dans la tâche était plus élevée chez les personnes autistes spécifiquement, ce qui constitue un marqueur potentiel pour dépister, et éventuellement identifier le TSA, bien que d’autres études soient nécessaires pour le confirmer. Cette étude novatrice a permis d’illustrer la complexité de la réalité clinique et de mettre la table pour des études subséquentes en psychiatrie computationnelle auprès des enfants. / Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder defined by two main diagnostic criteria: atypical social communication and restricted, repetitive interests including behavioral rigidity. Atypical social communication is typically measured and quantified by deficits in Theory of Mind (ToM). Conventional tools for assessing ToM have several limitations, including subjectivity bias, poor ecological validity and lack of reciprocal exchange. A dyadic, integrative task could be used to overcome these limitations. Such a task was tested in an adult autistic population. Specifically, the Penny-Hiding Game (PHG), in which a player must guess in which of his or her hands an opponent has hidden a coin, was administered to a group of autistic and a group of non-autistic individuals. The task was comprised of two framings: one social, where the participant believed he or she was playing online against a human being, and one non-social, where the participant was informed that he or she was playing against a computer. This study showed that autistic adults were insensitive to the framing in their behavior, which was modeled by a computational phenotype made up of two indices: sophistication and flexibility. In addition, the study identified an adaptation strategy that appeared to be specific to the autistic group. The computational phenotype also made it possible to classify participants with a specificity equivalent to standard diagnostic tools. The present study therefore aims to extend this understanding of social reciprocity in people with autism to a pediatric population. The issue of comorbidities in autism is also a major obstacle to scientific progress. We therefore decided to include in our sample two groups with the most prevalent comorbidities in autism: Attention-Deficit with or without Hyperactivity Disorder (ADHD) and Anxiety Disorder (AD). Following an analysis of performance with these three groups and visual inspection of choice sequences during the task, an alternation/perseveration parameter and a neurotypical (NT) group were added. These two additions make this master’s thesis partially exploratory. The aims were therefore to validate the PHG computerized task with clinical pediatric populations, to identify and explore group differences in adaptation strategies, performance and computational phenotype, and to explore the potential role of associations between computational phenotype, performance and symptoms associated with the three clinical conditions included in the study: autistic, attentional/executive and anxious symptoms. Analyses revealed that the NT group performed significantly better than the ASD, ADHD and AD groups, and that symptom quantity and severity were negatively associated with task performance score in the social framing. The most frequently used strategy across all groups and regardless of framing was reinforcement learning, and no differences in sophistication or flexibility were detected. These two results represent a lack of replication of previous work with autistic and NT adults. Perseverance in the task was higher in autistic individuals, which could be a potential marker for screening - possibly identifying - ASD, although further studies are needed to confirm this. This novel study helped illustrate the complexity of clinical contexts and set the table for subsequent computational psychiatry studies with children.