| 1. |
-
Cognitive and affective factors in relations to learning
- 2022
-
Samlingsverk (redaktörskap) (refereegranskat)abstract
- Both domain-general (e.g., working memory, executive functions) and domain-specific (e.g., number processing, phonological processing) cognitive factors have been found to predict learning in different age groups. Likewise, research has shown that various affective factors, such as different emotions (e.g., anxiety), self-concept, and interest, need to be considered when investigating individual differences in learning. However, few studies have investigated both cognitive and affective factors simultaneously in relation to learning. In particular, there is a lack of studies investigating the interplay (i.e., moderation and mediation) between cognitive and affective factors on learning.The goal of this Research Topic is to deepen our knowledge on the relations between learning and both cognitive and affective factors in different age groups. We aim to provide a broad scope of emerging areas in research on cognitive and affective factors, especially related to academic learning (e.g., mathematics, reading, and other school subjects). Studies focusing simultaneously looking at the interplay of these constructs, as well as longitudinally, are of great interest. Further, we are interested in innovative study designs and recent advances in methodology in this field. To promote quality education for all and equity in education, cognitive and affective factors related to aspects of learning ranging from pre-school to tertiary provision, and inclusion of individuals with special educational needs, are of interest.
|
|
| 2. |
|
|
| 3. |
- Eklöf, Hanna, 1974-, et al.
(författare)
-
Stress och påverkan på de nationella provresultaten för åk 3
- 2014
-
Konferensbidrag (refereegranskat)abstract
- Hur upplever 9-10 åringar de nationella proven i åk 3? Påverkar upplevelsen prestationen? Påverkas något matematikdelområde mer eller mindre av upplevelsen? Skiljer sig olika uppgifter åt beroende på om de har mer eller mindre text respektive bilder? Vad kan man som lärare tänka på och göra i allmänhet och i synnerhet vid prov/förhörssituationer?Läsåret 2012/13 genomfördes en studie med 624 st elever i åk 3 för att bringa klarhet i ovan frågor. Eleverna fick göra olika arbetsminnesövningar och svara på frågor om stress, motivation och attityder, etc. Teoretiskt tror man nämligen att allt för hög nivå av t ex stress (prestationsångest) sänker ens prestation på ett prov/förhör. Resultaten på de olika nationella delproven i matematik kördes därför statistiskt mot nivå av självrapporterad stress/ångest och uppskattad eller egentlig prestationsförmåga hos eleverna.Uppskattad eller egentlig förmåga att prestera i matematik för elever kan mätas genom t ex deras arbetsminneskapacitet. Arbetsminne är en kognitiv förmåga som är väl klarlagd för att väsentligen påverka prestation och utveckling inom t ex matematik- och läsförståelse hos både vuxna och barn (Menon, 2010). Det finns dessutom starka kopplingar mellan arbetsminneskapacitet och skolprestation i teoretiska ämnen. Majoriteten av de elever som har inlärningssvårigheter i skolan verkar även ha svag arbetsminnesförmåga (Gathercole et al., 2006).Arbetsminnet kan förenklat beskrivas som bestå av tre olika specialiserade komponenter. En huvudcentral som t ex kontrollerar, fördelar, uppmärksammar och processerar information, och hämtar/lagrar information från/i långtidsminnet. Till sin hjälp har denna huvudcentral en visuell-spatial del för hantering av bilder, former och dimension, samt en auditiv del för behandling av lingvistik (Baddeley, 1986). Matematik innefattar olika områden som beror av olika kognitiva förmågor (t ex huvudräkning, problemlösning), vilka i sin tur är relaterade till visuell-spatial och/eller auditiv fakta (Rasmussen & Bisanz, 2005).”Provstress” eller ”provängslan” är en etablerad term för att beskriva elevers påverkan och upplevelse av prov. Termen innefattar ofta för barn observerbara beteenden (t ex gå på toan, vicka på stolen, titta sig omkring), tankar/oro (t ex jag kommer aldrig att klara det här, mina föräldrar kommer att bli arga om jag misslyckas), autonoma/somatiska reaktioner (t ex svettas, ont i magen, varm om kinderna) (Zeidner, 2007). Man tror att provängslan är ett inlärt beteende som väcks tidigt i skolåren (Pekrun, 2000). Det är ett väldigt inskränkande tillstånd (Rothman, 2004) som starkt kan begränsa elevers prestation i alla åldrar (Birenbaum & Gutvirtz, 1993). Även om ett visst mått av provängslan är nödvändigt för att öka fokus, motivation och förberedelse (Gregor, 2005), kan det i allt för höga nivåer negativt påverka en elevs prestation och resultat på ett prov (Zeidner, 2007), särskilt i matematik (e.g. Putwain, 2008). Om och hur stark den kognitiva störningen är av provängslan används alltså i vår studie som ett mått eller symptom på elevers ev. underprestation. Vi undersöker också om eleverna uppvisar mer eller mindre av beteenden, autonoma reaktioner eller tankar relaterat till provängslan. Slutligen summerar vi våra resultat mot undervisning och prov/förhörssituationer. En jämförelse kommer även att göras med Finländska och Kinesiska åk-3 elever.
|
|
| 4. |
- Finell, Jonatan, et al.
(författare)
-
Working Memory and Its Mediating Role on the Relationship of Math Anxiety and Math Performance: A Meta-Analysis
- 2022
-
Ingår i: Frontiers in Psychology. - : Frontiers Media S.A.. - 1664-1078. ; 12
-
Tidskriftsartikel (refereegranskat)abstract
- It is well established that math anxiety has a negative relationship with math performance (MP). A few theories have provided explanations for this relationship. One of them, the Attentional Control Theory (ACT), suggests that anxiety can negatively impact the attentional control system and increase one's attention to threat-related stimuli. Within the ACT framework, the math anxiety (MA)—working memory (WM) relationship is argued to be critical for math performance. The present meta-analyses provides insights into the mechanisms of the MA—MP relation and the mediating role of WM. Through database searches with pre-determined search strings, 1,346 unique articles were identified. After excluding non-relevant studies, data from 57 studies and 150 effect sizes were used for investigating the MA—MP correlation using a random-effects model. This resulted in a mean correlation of r = −0.168. The database search of WM as a mediator for the MA—MP relation revealed 15 effects sizes leading to a descriptive rather than a generalizable statistic, with a mean indirect effect size of −0.092. Overall, the results confirm the ACT theory, WM does play a significant role in the MA—MP relationship.
|
|
| 5. |
- Flechard, Chris R., et al.
(författare)
-
Carbon-nitrogen interactions in European forests and semi-natural vegetation - Part 1: Fluxes and budgets of carbon, nitrogen and greenhouse gases from ecosystem monitoring and modelling
- 2020
-
Ingår i: Biogeosciences. - : Copernicus GmbH. - 1726-4170 .- 1726-4189. ; 17:6, s. 1583-1620
-
Tidskriftsartikel (refereegranskat)abstract
- The impact of atmospheric reactive nitrogen (N-r) deposition on carbon (C) sequestration in soils and biomass of unfertilized, natural, semi-natural and forest ecosystems has been much debated. Many previous results of this dC/dN response were based on changes in carbon stocks from periodical soil and ecosystem inventories, associated with estimates of N-r deposition obtained from large-scale chemical transport models. This study and a companion paper (Flechard et al., 2020) strive to reduce uncertainties of N effects on C sequestration by linking multi-annual gross and net ecosystem productivity estimates from 40 eddy covariance flux towers across Europe to local measurement-based estimates of dry and wet N-r deposition from a dedicated collocated monitoring network. To identify possible ecological drivers and processes affecting the interplay between C and N-r inputs and losses, these data were also combined with in situ flux measurements of NO, N2O and CH4 fluxes; soil NO3- leaching sampling; and results of soil incubation experiments for N and greenhouse gas (GHG) emissions, as well as surveys of available data from online databases and from the literature, together with forest ecosystem (BAS-FOR) modelling. Multi-year averages of net ecosystem productivity (NEP) in forests ranged from -70 to 826 gCm(-2) yr(-1) at total wet + dry inorganic N-r deposition rates (N-dep) of 0.3 to 4.3 gNm(-2) yr(-1) and from -4 to 361 g Cm-2 yr(-1) at N-dep rates of 0.1 to 3.1 gNm(-2) yr(-1) in short semi-natural vegetation (moorlands, wetlands and unfertilized extensively managed grasslands). The GHG budgets of the forests were strongly dominated by CO2 exchange, while CH4 and N2O exchange comprised a larger proportion of the GHG balance in short semi-natural vegetation. Uncertainties in elemental budgets were much larger for nitrogen than carbon, especially at sites with elevated N-dep where N-r leaching losses were also very large, and compounded by the lack of reliable data on organic nitrogen and N-2 losses by denitrification. Nitrogen losses in the form of NO, N2O and especially NO3- were on average 27%(range 6 %-54 %) of N-dep at sites with N-dep < 1 gNm(-2) yr(-1) versus 65% (range 35 %-85 %) for N-dep > 3 gNm(-2) yr(-1). Such large levels of N-r loss likely indicate that different stages of N saturation occurred at a number of sites. The joint analysis of the C and N budgets provided further hints that N saturation could be detected in altered patterns of forest growth. Net ecosystem productivity increased with N-r deposition up to 2-2.5 gNm(-2) yr(-1), with large scatter associated with a wide range in carbon sequestration efficiency (CSE, defined as the NEP/GPP ratio). At elevated N-dep levels (> 2.5 gNm(-2) yr(-1)), where inorganic N-r losses were also increasingly large, NEP levelled off and then decreased. The apparent increase in NEP at low to intermediate N-dep levels was partly the result of geographical cross-correlations between N-dep and climate, indicating that the actual mean dC/dN response at individual sites was significantly lower than would be suggested by a simple, straightforward regression of NEP vs. N-dep.
|
|
| 6. |
- Garcia, Johan, 1970-, et al.
(författare)
-
DIOPT : Extremely Fast Classification Using Lookups and Optimal Feature Discretization
- 2020
-
Ingår i: 2020 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN). - : IEEE. - 9781728169262
-
Konferensbidrag (refereegranskat)abstract
- For low dimensional classification problems we propose the novel DIOPT approach which considers the construction of a discretized feature space. Predictions for all cells in this space are obtained by means of a reference classifier and the class labels are stored in a lookup table generated by enumerating the complete space. This then leads to extremely high classification throughput as inference consists only of discretizing the relevant features and reading the class label from the lookup table index corresponding to the concatenation of the discretized feature bin indices. Since the size of the lookup table is limited due to memory constraints, the selection of optimal features and their respective discretization levels is paramount. We propose a particular supervised discretization approach striving to achieve maximal class separation of the discretized features, and further employ a purpose-built memetic algorithm to search towards the optimal selection of features and discretization levels. The inference run time and classification accuracy of DIOPT is compared to benchmark random forest and decision tree classifiers in several publicly available data sets. Orders of magnitude improvements are recorded in classification runtime with insignificant or modest degradation in classification accuracy for many of the evaluated binary classification tasks.
|
|
| 7. |
- Garcia, Johan, 1970-, et al.
(författare)
-
Efficient Distribution-Derived Features for High-Speed Encrypted Flow Classification
- 2018
-
Ingår i: NetAI'18 Proceedings of the 2018 Workshop on Network Meets AI & ML. - New York : Association for Computing Machinery (ACM). - 9781450359115 ; , s. 21-27
-
Konferensbidrag (refereegranskat)abstract
- Flow classification is an important tool to enable efficient network resource usage, support traffic engineering, and aid QoS mechanisms. As traffic is increasingly becoming encrypted by default, flow classification is turning towards the use of machine learning methods employing features that are also available for encrypted traffic. In this work we evaluate flow features that capture the distributional properties of in-flow per-packet metrics such as packet size and inter-arrival time. The characteristics of such distributions are often captured with general statistical measures such as standard deviation, variance, etc. We instead propose a Kolmogorov-Smirnov discretization (KSD) algorithm to perform histogram bin construction based on the distributional properties observed in the data. This allows for a richer, histogram based, representation which also requires less resources for feature computation than higher order statistical moments. A comprehensive evaluation using synthetic data from Gaussian and Beta mixtures show that the KSD approach provides Jensen-Shannon distance results surpassing those of uniform binning and probabilistic binning. An empirical evaluation using live traffic traces from a cellular network further shows that when coupled with a random forest classifier the KSD-constructed features improve classification performance compared to general statistical features based on higher order moments, or alternative bin placement approaches.
|
|
| 8. |
- Garcia, Johan, 1970-, et al.
(författare)
-
On Runtime and Classification Performance of the Discretize-Optimize (DISCO) Classification Approach
- 2018
-
Ingår i: Performance Evaluation Review. - New york, USA : Association for Computing Machinery (ACM). - 0163-5999 .- 1557-9484. ; 46:3, s. 167-170
-
Tidskriftsartikel (refereegranskat)abstract
- Using machine learning in high-speed networks for tasks such as flow classification typically requires either very resource efficient classification approaches, large amounts of computational resources, or specialized hardware. Here we provide a sketch of the discretize-optimize (DISCO) approach which can construct an extremely efficient classifier for low dimensional problems by combining feature selection, efficient discretization, novel bin placement, and lookup. As feature selection and discretization parameters are crucial, appropriate combinatorial optimization is an important aspect of the approach. A performance evaluation is performed for a YouTube classification task using a cellular traffic data set. The initial evaluation results show that the DISCO approach can move the Pareto boundary in the classification performance versus runtime trade-off by up to an order of magnitude compared to runtime optimized random forest and decision tree classifiers.
|
|
| 9. |
- Garcia, Johan, 1970-, et al.
(författare)
-
Towards Video Flow Classification at a Million Encrypted Flows Per Second
- 2018
-
Ingår i: Proceedings of 32nd International Conference on Advanced Information Networking and Applications (AINA). - Krakow : IEEE. - 9781538621967 - 9781538621950
-
Konferensbidrag (refereegranskat)abstract
- As end-to-end encryption on the Internet is becoming more prevalent, techniques such as deep packet inspection (DPI) can no longer be expected to be able to classify traffic. In many cellular networks a large fraction of all traffic is video traffic, and being able to divide flows in the network into video and non-video can provide considerable traffic engineering benefits. In this study we examine machine learning based flow classification using features that are available also for encrypted flows. Using a data set of several several billion packets from a live cellular network we examine the obtainable classification performance for two different ensemble-based classifiers. Further, we contrast the classification performance of a statistical-based feature set with a less computationally demanding alternate feature set. To also examine the runtime aspects of the problem, we export the trained models and use a tailor-made C implementation to evaluate the runtime performance. The results quantify the trade-off between classification and runtime performance, and show that up to 1 million classifications per second can be achieved for a single core. Considering that only the subset of flows reaching some minimum flow length will need to be classified, the results are promising with regards to deployment also in scenarios with very high flow arrival rates.
|
|
| 10. |
|
|
