The resources required to have humans score extended written response items in English language learner (ELL) contexts has caused automated essay scoring (AES) to emerge as a desired alternative. However, these systems often rely heavily on indirect proxies of writing quality such as word, sentence, and essay lengths because of their strong correlation to scores (Vajjala, 2017). This has led to concern about the validity of the features used to establish the predictive accuracy of AES systems (Attali, 2007; Weigle, 2013). Reliance on construct-irrelevant features in ELL contexts also forfeits the opportunity to provide meaningful diagnostic feedback to test-takers or provide the second language acquisition (SLA) field with real insights (C.-F. E. Chen & Cheng, 2008). This thesis seeks to improve the validity and reliability of an AES system developed for ELL essays by employing a new set of features based on the acquisition order of English negation. Modest improvements were made to a baseline AES system's accuracy, showing the possibility and importance of engineering features relevant to the construct being assessed in ELL essays. In addition to these findings, a novel ordering of the sequence of English negation acquisition not previously described in SLA research emerged.

<p>This work consists of two parts: the theoretical and the experimental. In the theoretical part, some general and some language specific theories of tense, aspect and aktionsart are presented, and the temporal systems of Swedish and Polish are compared. </p><p>The theoretical part is not a mere review of the literature on the subject. The comparison of the descriptions of aspect and aktionsart by Slavic researchers with the universal theory of Smith (1991) and (1977) and with description of aktionsart in Swedish in Teleman et al. (1999) has allowed me for some important observations as to the nature of the long-lasting dispute about the differences between aspect and aktionsart.</p><p>The experimental part is a cross-sectional study on the role of the linguistic context on the acquisition of the pluperfect by Polish learners Swedish as a foreign language. The informants are university students studying Swedish as a foreign language. The language samples were collected by means of two types of tests: gap-filling and translation from Polish. </p><p>Twelve linguistic factors, each divided into two subgroups, were hypothesised to have affected the correct use of the pluperfect. All those hypotheses as to which of the subgroups may inhibit and which may facilitate the correct use of the pluperfect are grounded in linguistic theories, i.e. presented in the form of linguistically-based discussions as to what kind of effect, facilitative or inhibiting, each of the linguistic factors may have had, and why. </p><p>The effect of those factors upon the correct use of the pluperfect has been tested by means of a step-wise multiple regression which measured the simultaneous effect of each factor upon the correct use of the pluperfect. This method has confirmed the facilitative effect of the following six linguistic factors: intrasentential indication of topic time (subordinate clause), unbounded verb indicating topic time, agentive meaning of the target verb, specifying subordinate clause, statal pluperfect and location of the time of action of pluperfect clause outside the temporal frame of narrative plot.</p>

This work consists of two parts: the theoretical and the experimental. In the theoretical part, some general and some language specific theories of tense, aspect and aktionsart are presented, and the temporal systems of Swedish and Polish are compared. The theoretical part is not a mere review of the literature on the subject. The comparison of the descriptions of aspect and aktionsart by Slavic researchers with the universal theory of Smith (1991) and (1977) and with description of aktionsart in Swedish in Teleman et al. (1999) has allowed me for some important observations as to the nature of the long-lasting dispute about the differences between aspect and aktionsart. The experimental part is a cross-sectional study on the role of the linguistic context on the acquisition of the pluperfect by Polish learners Swedish as a foreign language. The informants are university students studying Swedish as a foreign language. The language samples were collected by means of two types of tests: gap-filling and translation from Polish. Twelve linguistic factors, each divided into two subgroups, were hypothesised to have affected the correct use of the pluperfect. All those hypotheses as to which of the subgroups may inhibit and which may facilitate the correct use of the pluperfect are grounded in linguistic theories, i.e. presented in the form of linguistically-based discussions as to what kind of effect, facilitative or inhibiting, each of the linguistic factors may have had, and why. The effect of those factors upon the correct use of the pluperfect has been tested by means of a step-wise multiple regression which measured the simultaneous effect of each factor upon the correct use of the pluperfect. This method has confirmed the facilitative effect of the following six linguistic factors: intrasentential indication of topic time (subordinate clause), unbounded verb indicating topic time, agentive meaning of the target verb, specifying subordinate clause, statal pluperfect and location of the time of action of pluperfect clause outside the temporal frame of narrative plot.