Rabia Choudhery[Bearbeiten]

Eberhard Karls Universitaet Tuebingen
rabia-naeem.choudhery@student.uni-tuebingen.de
WS 2010/2011
7te Semester
Hauptfach: Psychologie
Kurs: CSCL

Warum sind Wikis gut zum lernen?

Wiki ist eine Webplatform wo Information gleichseitig an vielen geteil werden kann. Fuer formulierung von neues Wissen, kann dieser Information durch assimilation & akkomodation fuer individuellen und fuer gruppen hilfreich sein.

Notes from 25.10

-Web 2.0 -> blogs etc.Web as platform: platform independent(works with all devices such as computers, cell-phones, netbooks).
-Wikis -> websites linked, make changes or delete, collaborative artifact.
-Long trial curve shows that a few users use Wiki a lot and a large number of users make very few entries (90% are lurkers; people who read other's information but do not contribute anything themselves 9% users and 1% heavy users; keep up the site and do all the work).
-Two aspects of learning: 1. Individual Learning 2. Collaborative Knowledge Construction
-Theoretical Approaches:
--> knowledge building (Scardamalia & Bereiter, 2006): cooperative process in a community, a process where members are integrated and new knowledge occurs (when I produce new knowledge), ex: science and research, epistemic artifact for example a text book (the knowledge of a community is summed up) or a hands on artifact that I can grab and work with such a camera, a large interest in community knowledge.
--> cognitive conflict as a trigger (Piaget): learning occurs when a cognitive conflict occurs, conflict=learning is occurring, for example: child always says dog but at some point realizes that some say cat and learns the difference between a cat and a cat and other four legged animals, Piaget describes assimilation (take in new information or experiences and incorporate them into our existing ideas) and accommodation (altering one’s existing schemas, or ideas, as a result of new information or new experiences. New schemas may also be developed during this process.) as an individual processes.
--> co-evolution model (Cress & Kimmerle, 2008): assimilation and accommodation can occur for collaborative knowledge construction, but it also looks at individual process (imagine a + where on the right hand side we have collaborative learning and on the left hand side we have individual learning and at the top and bottom we have assimilation and accommodation). If the difference between two information sources (collaborative knowledge source and individual knowledge source) too high or too low then learning will not occur. Learning will only occur if the incongruousness is at a middle-level.

Presentation at ICSL 2008 "Learning and Knowledge building with wikis: The Impact of Incongruity between People’s Knowledge and a Wiki’s Information." about a study using the co-evolution model as theoretical basis.

Notes from 08.11

-worked in group on the expert text (Text Comprehension, Zwaan & Singer) and make changes to the pre-existing site "Construction-Integration-Modell" on wiki

-Construction-Integration-Model: the theory behind the model: texts are comprehended in chunks (a chunk is a clause or a sentence). The processing of each chunk includes a construction phase and an integration phase.

1.Construction phase: idea units are formed and called propositions. In this process both spelling and grammatical syntax of the text is analyzed. Propositions consist of a predicate and argument(s). The various propositions are then organized in a "coherent network." The structure of the network may contain links between the various propositions, conclusions which contribute the context of the text and text generalization.

2.Integration phase: activation of appropriate and inactivation of inappropriate pre-existing concepts. Deactivation creates a coherent text representation, which does not contain conflicting information or unnecessary propositions and the activated original network changes to long-term memory representations.Long-term memory encoding has three levels of representation of the comprehended text: 1. surface structure 2. text base 3. situation model

-Long-term memory representation
1. Surface structure: mental representation of each word of a sentence in its exact grammatical form and order. The storage of words depends on the attention that is given during the reading of the words. For example, the surface structure for a joke, an insult, songs and poems may be encoded better than the surface structure of a newspaper.

2. Text base: real semantic information /content of the text. Propositions are mainly stored when they are particularly relevant, or have been recently read.

3. Situation model: individual reorganization of an event in the text and reduction of semantic representation of the text based on prior knowledge and therefore, inferences may also be drawn. These can be defined by time, space, characters, causal and motivational context. The greater the consensus on the dimensions, the easier it is to integrate the events (the more similar the situation in two sentences, the more likely they are to be confused. This makes the situation model easier to be preserved in long-term memory.

-Coherence and Inference
1.Coherence: the identify relation between text ideas that exists between sentences or paragraphs. To identify this relation, co-referencing and anaphoric resolution is necessary.

1.1.Co-referencing: different linguistic expressions (synonyms or category relations) for a noun.
1.2.Anaphora: noun phrases/pronouns. Statements that refer to a person/object mentioned already in the text. Most anaphora are represented by pronouns. For example: the girl was holding the cup of tea in her hand and drank from it. Here “it” is the reference pronoun for the cup. Anaphoric processing requires identifying the referent (the cup) and computing the new information (“it” was drunken from) signaled by the anaphora (“it”).

2.Inference: help to identify the intended meaning of ambiguous words. Inferences can also be helpful for bridging the situation when the situation is spread across text idea (bridging inference).

2.1.Bridging inferences: the identification of causal relationships between sentences leads to comprehension. Bridging inferences may require more time when sentences/text ideas have more causal distance.

2.2.Elaborative inferences: text idea that are strongly implied on discourse but do not bear on coherence. They are necessary when the sentences are not necessarily connected by causal relationships with each other but rather belong to the same context. An elaborative inference allows the conclusion that a police officer was present when one reads that “a burglar was arrested”. There is conflicting evidence whether elaborative inferences accompany comprehension. Supporting evidence is that people appear to 1. encode implied semantic features, 2. elaborate category terms to their specific members 3. not reliably draw elaborative inferences about implied roles and, 4. draw elaborative inferences about text themes.

-Memory for Text

1.Question answering: serial processing of questions when the questions start with “what” and parallel when the questions start with “who”. Serial processing of questions is believed to progress with the following steps:
1.1.Encoding: similar to the construction phase where idea units called propositions are formed
1.2.Categorizing: requires syntactic and semantic analysis of the question when the relation of cause (why), reason (how) and time (when) are questioned 1.3.Strategy selection: two options are to either retrieve the question statement from memory or judge its plausibility. The guiding factors of strategy selection can be extrinsic (task instruction) or intrinsic (current activation). People can change their strategy for answering a question on the basis of advice and whether the question represents a recent or much earlier story.
1.4.Memory search: different procedures are in place for different category of questions.

2.Recall and recognition of text: recall of a text decreases systematically with time delay. Recognition of a text is influenced by the relation between the text and the test sentence. Recall responses can be classified as reproductions or reconstructions:
2.1.Reproductions: propositions expressed directly in the original article
2.2.Reconstructions: sensible guesses which do not have counterpart propositions in the text. Many recall responses of the situation model are reconstructions (because the surface structure and text base are forgotten and inferences are accepted as they are consistent with the situation model.

Notes from 15.11

-worked in a group on the Motivational Approach (Social Behavior) for Collaborative Learning (Text: Learning from Peers:Beyond he Rhetoric of Positive Results ro Angela M. O'Donnell & James O'Kelly) and made changes to the pre-existing wiki site "Motivational Approach"
-other groups worked on other approaches and made changes on other sites. Other approaches were Group Cohesion (Social Behavior), Vgotsky (Cognitive Approach), Piaget (Cognitive Approach), Elaboration

Notes from 22.11

-worked in expert groups and decided on how to introduce our texts next week
-made further changes to the topic "Construction-Integration-Modell" on wiki

Notes from 29.11

-came together in stem groups and introduced our texts as experts to others who were experts in other texts
-all the information on different topics can be found here: [Grundlagen]
-filled out the questionnaire based on the information learned from different groups on the following site: STAD-Test

Notes from 06.12

-went over the results of the STAD-Test in the seminar as a big group

Notes from 13.12

Notes from 20.12

Notes from 10.01

Einsatz von Medien
-frage:computer fuehrt zu mehr effizien?
-aber nicht nur technik, sondern: eingebettet in den konetxt!
-kultur beeinflusst den umfang mit technik

technike und kulturelle aspekte laut die autoren
-technik aspekte: schenller geht, nicht aufwendig, weniger psychologische bandbreite
-kulturelle aspekte: nutzer haben keine erfhrung, kein ettiquette (diese artikel war in 1984 geschrieben: ist das heute veraendert?)

sozialpsychologische aspekte
-zeit-/informationsverarbeitungs-druck
-kein regulierende feedback
-weniger ausdrucksmoeglichkeiten
-keine hinweise auf status/position
-anonymitaet
-fehlende normen/etiquette

zwei zentrale fragen
-mangel an kontext information
--->gut, wen gleichberectigter infofluss
--->schlecht, wenn problem bei der organisation
-wenig geteilte normen

-hat das sich heute veraendert?

3 studien
-studie: chat, chat+skript, e-mail
-at the time of the study, they didn't have internet, only local network, and only included texts, no images, attachments etc.

ergebnisse:effizienz
-problem: tippen dauert!
-CMC laenger als F"F bis einigung
--->schwerigkeit vehemenz einzuschaetzen
--->auch wenn zeit dauer tippen kontrolliert wird (da es mehr moeglichkeiten von non-verbales information zu teilen die mit computer nicht machbar war)
-weniger aeusserungen in CMC
-kein unterschied in der aufgabenorientierung

ergebnisse: partizipation
-sind alle gleich maessig auf die kommunikation geteiligt?
-verteilung aeusserungen/mitglieder

ergebnisse: group choice
-wie schnell einig sie all auf der grupe order einigen sie nicht?
-CMC hoeher choice shift
--->depersonalisierung?
--->weniger oziale druck sich zu einigen

ergebnisse: uninhibitation
-mehr bei CMC
-hoeher bei conference als bei mail

erklaerung

wir machen ein experiment
-UV: chatten/mail/F2F
-AV:?--->muessen selber eitscheiden was beobachten wir?