According to the theory of desirable difficulty, conflict during practice can elicit a greater degree of processing and result in a later memory benefit (Bjork & Bjork, 1994). The present paper extends the work by Thomson et al. (in prep) that there may be a stage-specific mechanism involved. They found a desirable difficulty effect when directing a conflict towards the categorization stage through classifying names by gender. However, no such effect was seen when classifying words by size. They provided evidence that gender classification is more semantically central than categorizing items by size, which tends to be relative and depend on context. We took the same stimuli used in the Thomson et al. (in prep) but had participants make animacy judgments (animal or “thing”) on words. A subsequent memory test revealed a desirable difficulty effect for incongruent words compared to congruent. Interestingly, animal words were better remembered overall compared to “thing” words. A second experiment directed a conflict towards the categorization and response processing stages through classifying names as male and female with semantic (male/female) and response selection (left/right) primes. A subsequent memory test revealed a desirable difficulty pattern of results (although non-significant) where incongruent compared to congruent words were better remembered for the sematic primes and congruent compared to incongruent words were better remembered for response primes. These results suggest to-be-remembered material needs to be the focus of attention and increasing difficulty to any stage of processing does not give you a guaranteed desirable effect. / Thesis / Master of Science (MSc)

This thesis investigates predictions from prominent conflict theories of cognitive control that information experienced under high conflict conditions should be better encoded. More specifically, recent research suggests that selectively attending to relevant stimuli while ignoring conflicting stimuli can lead to better memory. These ideas have been broadly discussed in the desirable difficulty literature – described by instances where increasing difficulty during initial task performance leads to better later memory. As a growing number of studies have attempted to produce these effects with mixed success, calls for more focused investigations into the underlying mechanisms have been made. This encoding benefit for high-control-demand or high-difficulty situations has been broadly conceptualized as a task-general property, where all activated representations should be better encoded. The goal of this thesis was to investigate whether memory-enhancing effects of difficulty manipulations depend on inducing additional cognitive control at particular information processing stages. This thesis documents some of the first work showing that the within-task locus of conflict and attentional control is critical to whether later memory benefits are seen – conflict/control focused on semantic item representation produces better memory, but conflict/control focused away from item representations at response selection gives no memory benefit. These findings and theory are then extended to physiological measures of pupil dilation and sequential (Grattron-like) conflict/control situations. This thesis proposes a stage-specific conflict-encoding model which complements and extends current leading theories of conflict-driven cognitive control. / Dissertation / Doctor of Philosophy (PhD) / There is an intuitive notion that making a learning experience hard will hinder memory of that information later on. Contrary to this belief, in certain circumstances, making learning difficult can actually enhance the memory of that information – this has been termed desirable difficulty. The issue with these desirable difficulties is that they are only sometimes effective. Originally it was proposed that general task-wide difficulty would lead to an enhancement in memory. This thesis, however, provides evidence suggesting that task difficulty is stage-specific in nature, meaning that for the difficulty to enhance memory, the difficulty needs to be at a specific stage of cognitive processing. For difficulty to have a beneficial effect on memory, the particular difficulty needs to focus an individual’s attention on the core meaning of what they are trying to remember, or else the difficulty will direct attention away from this important information causing a possible decrease in memory. These findings provide a framework for how and when to use difficulty as a means to enhance learning.

It is well established that repeated testing is more beneficial for durable learning than repeated studying of the same material, a phenomenon known as the testing effect. This study sought to investigate the role of working memory capacity (WMC) in relation to the learning process and the difficulty of the material to be learned when using a test-enhanced learning method. As between subject manipulation, participants (n = 99, M = 25.62 years of age) were divided into two groups, one using repeated studying and one using alternated testing and studying. A material of two difficulty levels, as well as immediate and delayed retention tests, was used in each condition as within subject manipulation. Further, an n-back task was used to measure WMC. Results from mixed model ANOVAs showed no significant impact of WMC on either the learning process or retention in relation to the difficulty of the material. The testing condition performed significantly higher than the studying condition on the retention tests. The testing effect is further cemented as a promising method for practical application in the educational sector regardless of both WMC and difficulty level. / Det är väl etablerat att upprepad testning är mer fördelaktigt för hållbar inlärning än upprepad instudering av samma material, ett fenomen känt som testeffekten. Denna studie ämnade undersöka arbetsminnets roll i relation till inlärningsprocessen och svårighetsgrad av material med testbaserat lärande som metod. Som mellangruppsmanipulation delades deltagare (n = 99, M = 25.62 år gamla) in i två grupper, en som upprepade gånger studerade materialet och en som alternerade studerande med tester. Ett material med två svårighetsgrader och ett direkt samt fördröjda retentionstester användes som inomgruppsmanipulation. Vidare användes ett n-backtest som mått på arbetsminneskapacitet. Resultat visade ingen signifikant inverkan av arbetsminne på varken inlärningsprocessen eller retention i relation till svårighetsgrad av material. Testbetingelsen presterade signifikant högre på retentionstest än studiebetingelsen. Testeffekten fastställs ytterligare som lovande metodik för praktisk applikation i utbildningssektorn oberoende av både arbetsminneskapacitet och svårighetsgrad.