This multi-method study provides detailed information about the friendships of 87 children with ADHD (77.0% boys) and 46 comparison children (73.9% boys) between the ages of 7 and 13. The methods used in the study included parent and teacher ratings, self-report measures, and direct observation of friends’ dyadic behaviours in three structured analogue tasks. Results indicated that, in contrast with comparison children, children with ADHD had friends with high levels of ADHD and oppositional symptoms; they perceived fewer positive features and more negative features, and were less satisfied in their friendships. Observational data indicated that children with ADHD performed both more legal and more illegal manoeuvres than comparison children in a fast-paced competitive game. While negotiating with their friends, children with ADHD made more insensitive and self-centred proposals than comparison children. In dyads consisting of one child with ADHD and one typically developing child, children with ADHD were often more controlling than their non-diagnosed friends. Globally, these results were robust and did not seem to be affected by age differences, ADHD subtypes, comorbidities, and medication status. Given the increased recognition of ADHD in adolescence and adulthood as well as the fact that negative peer reputation in childhood very strongly predicts mental-health status by early adulthood, this research may lead to the discovery of meaningful ways to help people with ADHD achieve improved mental health and happiness over their lifespan.

This multi-method study provides detailed information about the friendships of 87 children with ADHD (77.0% boys) and 46 comparison children (73.9% boys) between the ages of 7 and 13. The methods used in the study included parent and teacher ratings, self-report measures, and direct observation of friends’ dyadic behaviours in three structured analogue tasks. Results indicated that, in contrast with comparison children, children with ADHD had friends with high levels of ADHD and oppositional symptoms; they perceived fewer positive features and more negative features, and were less satisfied in their friendships. Observational data indicated that children with ADHD performed both more legal and more illegal manoeuvres than comparison children in a fast-paced competitive game. While negotiating with their friends, children with ADHD made more insensitive and self-centred proposals than comparison children. In dyads consisting of one child with ADHD and one typically developing child, children with ADHD were often more controlling than their non-diagnosed friends. Globally, these results were robust and did not seem to be affected by age differences, ADHD subtypes, comorbidities, and medication status. Given the increased recognition of ADHD in adolescence and adulthood as well as the fact that negative peer reputation in childhood very strongly predicts mental-health status by early adulthood, this research may lead to the discovery of meaningful ways to help people with ADHD achieve improved mental health and happiness over their lifespan.

By using a three-choice preference test I tested food preferences in eight meerkats (Sutricata suricatta) for nine different food items. The meerkats were presented with all possible combinations of 9 food items, in total 84 combinations. The meerkats displayed the following rank order of food preference: cricket > zophoba lava > egg white > banana = tomato > orange > apple > cucumber > carrot. A correlation test between the food preference and nutritional content showed that the preference is significantly correlated with protein content. There is also a significant correlation between ash content and the food preference. There were no correlations with water, fat or vitamins and minerals. This suggests that, as in the wild, the captive meerkats at Kolmården wild life park seek food with high protein content. Ash has a high significant correlation with protein, which may explain the correlation between ash and the food preference. A comparison between the individual’s food preferences showed that the oldest individuals had the highest number of correlations with their food preferences to other individual’s food preference. This suggests an observational learning capability from older to younger individuals, which has also been reported in wild meerkat

This DPhil dissertation presents two new and independent samples of faint radio sources. The first sample is the 37 SXDS radio sources with flux densities at 1.4-GHz above 2 mJy, a spectroscopic completeness of 65% and a median redshift z_med ≈ 1.1. The second sample is the 47 TOOT00 radio sources with flux densities at 151 MHz above 100 mJy, a spectroscopic completeness of 85% and z_med ∼ 1.25. Optical, near- and mid-IR photometry, optical spectroscopy, and radio observations are used in the analysis and comparison of the two samples. The quasar fraction in the TOOT00 radio sources is 0.13 < f_q < 0.25 above the FRI/FRII break in radio luminosity, while use of 24 μm data reveals objects with significant but sometimes obscured accretion and gives quasar-mode fraction of 0.5 → 0.9 above the FRI/FRII break. The FRI/FRII divide seen at z < ∼ 0.5 is also observed at z ∼ 1 for FRII objects in the TOOT00 and SXDS samples, but examples of FRI radio sources above the FRI/FRII break do exist. The total number of the TOOT00 objects and their distribution are consistent with simulations based on extrapolations from previous work, while for the SXDS objects the results are only broadly similar. Based on that comparison, the redshift spikes seen at z ∼ 1.3 in TOOT00 and at z ∼ 0.65 & 2.7 in SXDS appear to be significant, and might be due to Large-Scale Structure. A V/Vmax test suggests the cosmic evolution of the TOOT00 and SXDS samples, is different. The TOOT00 radio sources are 2-times more luminous in host-galaxy starlight than the SXDS radio sources. The almost proportionality between radio luminosity at 1.4 GHz and 24 μm luminosity suggests that L_rad traces accretion luminosity and L[OII] ∝ L_rad^0.7 may reflect imperfections in the L[OII]-accretion luminosity scaling. Mid-IR 24 μm observations in the SXDS sample suggest that 30% of the light from the nucleus is absorbed by the torus and re-emitted in the mid-IR, while ∼ 1% of the light is scattered above and below the torus.

While the role of attributional feedback on student learning and achievement has been previously studied in laboratory settings, the importance of these types of feedback in real-time classroom settings is yet to be examined. This study attempted at exploring whether attributional feedback is present in interactions between teachers and students in natural classroom settings and how much they contribute to students' achievement in mathematics. Using an observational coding system, 55 one hour mathematics classrooms were coded for three different types of attributional feedback: direct attributional feedback, indirect attributional feedback, and strategy feedback. Direct attributional feedback consisted of explicit statements that were effort-oriented, ability- oriented, or knowing-oriented. Indirect attributional feedback consisted of teacher behavior that implicitly conveyed attributional messages to students. Behaviors such as unsolicited offers of help, giving credit to students, abandoning students, and calling out student name publicly were coded as indirect attributional behavior in this study. In addition, "why", "how", "what" strategy questions and strategy statements were coded as strategy feedback in this study. The results of this study indicated that while both direct and indirect attributional feedback are present in natural classroom settings, there is a significant difference between the number of times each type occurs. Teachers are more likely to convey attributional messages to students through indirect behavior compared to explicit attributional statements. When these types of feedback were examined in a model predicting students' mathematics achievement scores, the results showed that attributional feedback on their own significantly predicted student achievement but when it was combined with strategy feedback a stronger association with achievement scores occurred. Amazingly, it should be noted that in this model, direct attributional feedback indicated a negative association with mathematics achievement whereas indirect attributional feedback and strategy feedback indicated a positive association. These findings reveal that attributional feedback play an important role in student achievement in mathematics and the most effective way to present students with these types of feedback is to convey them indirectly and in combination with strategy feedback.

This thesis contributes to the research field of causal inference, where the effect of a treatment on an outcome is of interest is concerned. Many such effects cannot be estimated through randomised experiments. For example, the effect of higher education on future income needs to be estimated using observational data. In the estimation, assumptions are made to make individuals that get higher education comparable with those not getting higher education, to make the effect estimable. Another assumption often made in causal inference (both in randomised an nonrandomised studies) is that the treatment received by one individual has no effect on the outcome of others. If this assumption is not met, the meaning of the causal effect of the treatment may be unclear. In the first paper the effect of college choice on income is investigated using Swedish register data, by comparing graduates from old and new Swedish universities. A semiparametric method of estimation is used, thereby relaxing functional assumptions for the data. One assumption often made in causal inference in observational studies is that individuals in different treatment groups are comparable, given that a set of pretreatment variables have been adjusted for in the analysis. This so called unconfoundedness assumption is in principle not possible to test and, therefore, in the second paper we propose a Bayesian sensitivity analysis of the unconfoundedness assumption. This analysis is then performed on the results from the first paper. In the third paper of the thesis, we study profile likelihood as a tool for semiparametric estimation of a causal effect of a treatment. A semiparametric version of the Bayesian sensitivity analysis of the unconfoundedness assumption proposed in Paper II is also performed using profile likelihood. The last paper of the thesis is concerned with the estimation of direct and indirect causal effects of a treatment where interference between units is present, i.e., where the treatment of one individual affects the outcome of other individuals. We give unbiased estimators of these direct and indirect effects for situations where treatment probabilities vary between individuals. We also illustrate in a simulation study how direct and indirect causal effects can be estimated when treatment probabilities need to be estimated using background information on individuals.

The state-based approach plays an important role in modeling and control of Discrete-Event Systems (DES). Based on previous work, state feedback control of DES with nonblocking is thoroughly investigated; a general construction method for memory and the corresponding predicate is also specified. Two examples are provided in illustration. Also presented is state-based control of DES with observational abstraction. Based on the existing idea of quasi-congruence, quasi-observer, as a kind of observational abstraction, is developed; its advantages and limitations are discussed by means of simple examples. Imposing an observational partition on the state set also leads to observational abstraction. On this basis, the state-feedback controller design is introduced; in particular, the notion of high and low modeling levels for DES is proposed, based on which reachability and controllability are further discussed and compared. Finally, two simple applications are provided to show the advantage of observational partition in DES analysis and control.

The state-based approach plays an important role in modeling and control of Discrete-Event Systems (DES). Based on previous work, state feedback control of DES with nonblocking is thoroughly investigated; a general construction method for memory and the corresponding predicate is also specified. Two examples are provided in illustration. Also presented is state-based control of DES with observational abstraction. Based on the existing idea of quasi-congruence, quasi-observer, as a kind of observational abstraction, is developed; its advantages and limitations are discussed by means of simple examples. Imposing an observational partition on the state set also leads to observational abstraction. On this basis, the state-feedback controller design is introduced; in particular, the notion of high and low modeling levels for DES is proposed, based on which reachability and controllability are further discussed and compared. Finally, two simple applications are provided to show the advantage of observational partition in DES analysis and control.

This multi-method study provides detailed information about the friendships of 87 children with ADHD (77.0% boys) and 46 comparison children (73.9% boys) between the ages of 7 and 13. The methods used in the study included parent and teacher ratings, self-report measures, and direct observation of friends’ dyadic behaviours in three structured analogue tasks. Results indicated that, in contrast with comparison children, children with ADHD had friends with high levels of ADHD and oppositional symptoms; they perceived fewer positive features and more negative features, and were less satisfied in their friendships. Observational data indicated that children with ADHD performed both more legal and more illegal manoeuvres than comparison children in a fast-paced competitive game. While negotiating with their friends, children with ADHD made more insensitive and self-centred proposals than comparison children. In dyads consisting of one child with ADHD and one typically developing child, children with ADHD were often more controlling than their non-diagnosed friends. Globally, these results were robust and did not seem to be affected by age differences, ADHD subtypes, comorbidities, and medication status. Given the increased recognition of ADHD in adolescence and adulthood as well as the fact that negative peer reputation in childhood very strongly predicts mental-health status by early adulthood, this research may lead to the discovery of meaningful ways to help people with ADHD achieve improved mental health and happiness over their lifespan.

We have observed the nearest ultraluminous infrared galaxy Arp 220 with a 13 month near-infrared observing program using the Canada France Hawaii Telescope to search for obscured supernovae in this extreme star forming environment. This monitoring program was aimed as a feasibility study to determine the practicality of a large scale near-IR LIRG/ULIRG imaging survey. Establishing the supernova rate in these dusty galaxies is an important step toward confirming theorized star formation rates and settling the debate between the dominant energy source in LIRGs: star formation or AGN activity. Both the deduced high star formation rate and far-IR luminosity of Arp 220 suggest an atypically high supernova rate of 1-4 per year, two orders of magnitude greater than that of the Milky Way. We attempt the first direct observation of this rate which to date has been probed primarily through radio measurements of supernovae and remnants. Through a point-spread function matching and image subtraction procedure we find no supernovae outside the galactic nucleus, consistent with the paradigm of a strong nuclear-contained starburst. Image subtraction residuals prevent the discovery of supernovae in the central regions of the galaxy. Using differential photometry we find evidence for a statistically significant brightening in the Arp 220 nucleus with a K-band peak of approximately $\Delta m_K=0.16$ magnitudes. To find the true peak magnitude we use Hubble Space Telescope archival data to subtract off the nuclear background and find an absolute magnitude of $M_K = -22.19 \pm 0.16$ (non-absorbed). This exceeds the luminosity of a typical core collapse supernova by roughly 3.5 magnitudes; rather, the observed variations in nuclear brightness are most likely the signature of an active galactic nucleus embedded in the dusty nuclei of Arp 220 or the superposition of light from several supernovae. This method is not sensitive to the detection of individual supernovae and we cannot rule out the occurrence of any nuclear SNe during the observing period. The brightening event is dimmer in the H and J bands, appearing to be affected by extinction. Interpreting this as a supernova-related event we estimate the extinction in the nuclear regions of Arp 220 to lie between $2.01 \le A_K \le 3.40$ or $17.95 \le A_V \le 30.36$ in the optical, in agreement with several other estimates. Improved resolution is required in order to detect supernovae in the extremely bright nuclear environments of LIRGs. Alternatively, infrared spectroscopy would reveal the telltale spectral features of nuclear supernovae. Spectroscopic observations of the Arp 220 nuclei were conducted using Keck in July 2013 for this very purpose; results are pending. We also explore the hypothesis that type Ia supernovae are produced primarily from young stellar populations. We model elliptical galaxies as two component stellar systems using PEGASE stellar templates: a fixed older underlying population coupled with a younger, less massive population. Varying the age and mass ratio of the young component, we examine its effect on I) the colours and II) the supernova rate of the single underlying population. We explore the effect with redshift and employ both theoretical and observational forms of the type Ia delay-time distribution. We then apply our models to the MENeaCS supernova survey and find that the number and distribution of red sequence SN Ia hosts agrees with theoretical expectations. The lack of evidence for a type Ia rate cutoff argues for a continuous delay-time distribution in support of the double degenerate model as the primary SN Ia progenitor channel. We conclude that it is not possible for all type Ia events in ellipticals to originate from a young frosting of stars. / Graduate / 0606 / jkezwer@uvic.ca