experimental design proposal

10
Experimental Design Experimental Design Proposal Proposal COMM 268 COMM 268 Wuping Lu Wuping Lu

Upload: elton-bowman

Post on 31-Dec-2015

16 views

Category:

Documents


0 download

DESCRIPTION

Experimental Design Proposal. COMM 268 Wuping Lu. Variables - independent. Anonymity of input vs. identity on public display Accountability vs. no accountability message on private display. Variables - manipulation. Accountability. Variables - manipulation. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Experimental Design Proposal

Experimental Design ProposalExperimental Design Proposal

COMM 268COMM 268

Wuping LuWuping Lu

Page 2: Experimental Design Proposal

Variables - Variables - independentindependent

Anonymity of input vs. identity on Anonymity of input vs. identity on public displaypublic display

Accountability vs. no accountability Accountability vs. no accountability message on private displaymessage on private display

Page 3: Experimental Design Proposal

Variables - Variables - manipulationmanipulation

2 x 2 between subjects factorial design2 x 2 between subjects factorial design

IdentityIdentity

Anonymity of input Anonymity of input and accountability and accountability message message

Identity and Identity and accountability messageaccountability message

Anonymity of input Anonymity of input and no accountability and no accountability message message

Identity and no Identity and no accountability messageaccountability message

Accou

nta

bility

Page 4: Experimental Design Proposal

Variables - Variables - manipulationmanipulation

Can be within-subjects design, too.Can be within-subjects design, too.• Strength - statistical power increases Strength - statistical power increases

and the probability of beta error and the probability of beta error decreasesdecreases

• Weakness – carryover effect: either Weakness – carryover effect: either fatigue or practice effect fatigue or practice effect

Page 5: Experimental Design Proposal

Variables - Variables - dependentdependent Collaborative effectiveness was measured in Collaborative effectiveness was measured in

terms of participants’ contribution and their terms of participants’ contribution and their evaluation of their collaboration experience. evaluation of their collaboration experience. • The amount of ideas generated. The amount of ideas generated. • Self-reported collaboration experienceSelf-reported collaboration experience

Found the group work a good collaborative experienceFound the group work a good collaborative experience Motivated to do best workMotivated to do best work More confident in expressing ideasMore confident in expressing ideas Contributed more personally because of the Contributed more personally because of the

collaboration formatcollaboration format Participated activelyParticipated actively More involvedMore involved Achieved more as a group because of the collaboration Achieved more as a group because of the collaboration

formatformat Group work was fun Group work was fun

Page 6: Experimental Design Proposal

TaskTask Visiting the ldt website: Visiting the ldt website: http://http://

ldt.stanford.eduldt.stanford.edu// Discuss and critique the website in terms Discuss and critique the website in terms

of strength and weaknessof strength and weakness Based on their critique, brainstorm what Based on their critique, brainstorm what

the future ldt website should look like in the future ldt website should look like in terms of content, features, functions and terms of content, features, functions and organization (architecture) organization (architecture)

Generate a final report including the Generate a final report including the critique and new design (expectation) of critique and new design (expectation) of the website the website

Page 7: Experimental Design Proposal

RationalesRationales Group work or collaboration has potential Group work or collaboration has potential

gains and losses.gains and losses. Increasing process gains and reducing Increasing process gains and reducing

process losses can facilitate group process losses can facilitate group interactions and improve group interactions and improve group performance. performance.

Specific procedures (interface) can be built Specific procedures (interface) can be built into computer-based collaboration process into computer-based collaboration process to achieve this purpose. to achieve this purpose.

We can manipulate these procedures to We can manipulate these procedures to see their effects.see their effects.

Page 8: Experimental Design Proposal

RationalesRationalesGroup Process GainsGroup Process Gains Group Process LossesGroup Process Losses

A group as a whole generates A group as a whole generates more information and more information and alternatives compared to the alternatives compared to the average group memberaverage group member

Groups are more effective and Groups are more effective and objective in evaluation and objective in evaluation and error detection taskserror detection tasks

Working in a group may Working in a group may motivate the individual motivate the individual member to perform bettermember to perform betterInteractions among the group Interactions among the group members lead to synergiesmembers lead to synergies

Member participation in the group is Member participation in the group is fragmented (i.e., group members fragmented (i.e., group members should take turns in should take turns in speaking/controlling the public speaking/controlling the public display)display)

One or a few individual members may One or a few individual members may dominate group discussions and dominate group discussions and monopolize the group’s time (control)monopolize the group’s time (control)

Fear of negative evaluation Fear of negative evaluation (evaluation apprehension) causes (evaluation apprehension) causes members to withdraw and avoid members to withdraw and avoid participating in the group discussionsparticipating in the group discussions

Higher volumes of information Higher volumes of information generated during the group process generated during the group process creates a condition of information creates a condition of information overload for individual memberoverload for individual member

Page 9: Experimental Design Proposal

Hypothesis Hypothesis Anonymity of input on public display and Anonymity of input on public display and

accountability message on private display accountability message on private display enhances the effectiveness of collaboration enhances the effectiveness of collaboration (defined in terms of ideas generated and self-(defined in terms of ideas generated and self-reported collaboration experience) by increasing reported collaboration experience) by increasing group process gains and decreasing group group process gains and decreasing group process losses.process losses.• Anonymity of input can decrease or eliminate evaluation Anonymity of input can decrease or eliminate evaluation

apprehension (a process gain) leading to an increase in apprehension (a process gain) leading to an increase in member participation and amount of information member participation and amount of information generated by the group (a process gain)generated by the group (a process gain)

• Showing message “everyone has to make contribution” Showing message “everyone has to make contribution” on each private display can increase individual on each private display can increase individual accountability and thus member participation and accountability and thus member participation and amount of information generated by the group (process amount of information generated by the group (process gain)gain)

Page 10: Experimental Design Proposal

Thanks!Thanks!

Questions?Questions?