The Design and Evaluation of Multitouch Marking Menus: Lepinski1,Grossman2,Fitzmaurice, chi 2010
–there are 31 possible combinations of putting fingers on a surface and they are called chords. this paper investigates performing all 31 comments in eight possible directions (N, NW, W, and so on).
-'Angular accuracy was very good, with an average angular error of only 5.6 degrees (Figure 7). In total, 98.2% of the gestures were completed with an angular accuracy of less than the 22.5 degrees which would be needed to select between eight radial items'
-' Chords with “gaps”, that is, requiring non-consecutive fingers to be depressed, tended to produce higher articulation times.'
GestureBar: Improving the Approachability of Gesture-based Interfaces Andrew Bragdon, Robert Zeleznik, Brian Williamson†, Timothy Miller, Joseph J. LaViola Jr., CHI 2009
- an attempt on creating a gesture learning environment for a diagram tool similar to the ribbon in the word 2007.
- extensive literature review on learning and presenting gestures. A combination of icons, animations, and descriptions were found to be best for teaching novices.
Experimental Analysis of Touch-Screen Gesture Designs in
Mobile Environments: Andrew Bragdon1, Eugene Nelson1, Yang Li2 and Ken Hinckley3, CHI 2011
-investigates the impact of distraction (walking, observing an object, performing mouse movement) on different types of mobile phone interaction (free shape gestures, bezel gestures, on-screen buttons)
-' We find that in the presence of environmental distractions, ges- tures can offer significant performance gains and reduced attentional load, while performing as well as soft buttons when the user‟s attention is focused on the phone'
- situation awareness task was to hold space button when a circle is displayed on a screen. Attention saturation task was to follow a dot with a computer mouse
–to simulate expert users every time an icon representing the gesture was displayed
–Participants were motivated to perform to their quickly and accurately with money incentives. reward was increased for a low average time, and decreased for a high error rate.
- The bezel gestures were found to be consistently faster and more accurate
- This result is excellent, showing that gestures can reduce user‟s attentional load over soft buttons, even though gestures involve reproducing geometry with some degree of accuracy and soft buttons can be performed in ~1.2 s
Gestural and Audio Metaphors as a Means of Control for Mobile Devices: Antti Pirhonen, Stephen Brewster and Christopher Holguin, chi 2002
- users were interacting with the touchscreen music player while walking. The advice was mounted on their hip. NASA TLX (with added annoyance measure) was used to record a workload, and the deviation from preferred walking speed was used as a measure of effort.
- The learning curve diagram was used to analyse the results: Y axis is a success rate of the last five trials. It looks really informative!
Squeeze Me, Hold Me, Tilt Me! An Exploration of Manipulative User Interfaces, Beverly ‘8;.Harrison, ‘Kenneth P. Fishkin, Anuj Gujar, Carlos Mochon*, Roy Want: chi 1998
- an attempt to add to a mobile digital text reader natural interactions known from reading a book. Series of tilt and pressure sensors were used to imitate turning pages and opening a book in a certain spot in the middle.
Wednesday, 15 June 2011
Thursday, 2 June 2011
Papers 4 - more interaction and CLT
Paas et al 2003:Cognitive Load Measurement as a Means to Advance Cognitive Load Theory
–argues that the combination of performance and cognitive load measures has been identified to constitute a reliable estimate of the mental efficiency of instructional methods
- the overview of progress on measuring cognitive load is presented over the last 15 years. Researchers are moving towards less intrusive ways of measuring loads, like questioners and biosensing. While secondary task is a reliable to, it can interfere considerably with the primary task
–authors propose a measure of instructional efficiency which represents the relation between performance and mental effort.
- 'An interesting observation is that even though researchers are continuously trying to find or develop physiological and secondary task measures of cognitive load, subjective work- load measurement techniques using rating scales remain pop- ular, because they are easy to use; do not interfere with primary task performance; are inexpensive; can detect small variations in workload (i.e., sensitivity); are reliable; and pro- vide decent convergent, construct, and discriminate validity (Gimino, 2002; Paas et al., 1994). However, we must stress that the internal consistency of these measures requires fur- ther studies'
- others inverstigated split attention, expertie and age differences impact.
Oviatt 2006
-very well presented paper with exceptional graphs
-'the study evaluated whether student performance would deteriorate as interfaces departed more from students’ existing work practice (GT > PT > DP), with lower-performing students experiencing greater cognitive load and performance degradation than high-performing students when using the same interfaces. '
-'Typical performance measures of cognitive load have included time to complete tasks, reaction time, correct solutions, memory retrieval time and correctness, time estimation, rate of physical activity and speech, spoken disfluencies, multimodal integration patterns, and other indices'
- 'Dual-task methods are especially relevant and ecologically valid when applied to field and mobile interface design, which chronically involve multitasking and divided attention.'
- think aloud protocol was used to assess whether learners were working on a low level of high-level issue
Oviatt, COulston, Lonsford 2004:When Do We Interact Multimodally? Cognitive Load and Multimodal Communication Patterns
- 'one critical objective of all mobile interface design is the need to manage multitasking, interruption, attentional distraction, fluctuations in the difficulty of natural field tasks and situations, and resulting cognitive overload [8, 18]. This is essential because mobile users need to focus on complex primary field tasks that can vary substantially in difficulty and also involve dual tasking between the field task and secondary tasks involved in controlling an interface. '
- users were using multimodal pen and word interfaces, but findlings are interesting, more multimodal interactions were found when working in a familiar area and when task was more difficult. It is possible that users maintain the working memory resources and distribute tasks to more centres when the task becomes difficult.
Sweller, Merrienboer, Paas 1998:Cognitive Architecture and Instructional Design
- in this paper mechanisms concerning instructional procedures are described in depth: the goal free effect, worked example affect, completion problem effect, split attention effect modality effect, redundancy effect, variability affects,
–argues that the combination of performance and cognitive load measures has been identified to constitute a reliable estimate of the mental efficiency of instructional methods
- the overview of progress on measuring cognitive load is presented over the last 15 years. Researchers are moving towards less intrusive ways of measuring loads, like questioners and biosensing. While secondary task is a reliable to, it can interfere considerably with the primary task
–authors propose a measure of instructional efficiency which represents the relation between performance and mental effort.
- 'An interesting observation is that even though researchers are continuously trying to find or develop physiological and secondary task measures of cognitive load, subjective work- load measurement techniques using rating scales remain pop- ular, because they are easy to use; do not interfere with primary task performance; are inexpensive; can detect small variations in workload (i.e., sensitivity); are reliable; and pro- vide decent convergent, construct, and discriminate validity (Gimino, 2002; Paas et al., 1994). However, we must stress that the internal consistency of these measures requires fur- ther studies'
- others inverstigated split attention, expertie and age differences impact.
Oviatt 2006
-very well presented paper with exceptional graphs
-'the study evaluated whether student performance would deteriorate as interfaces departed more from students’ existing work practice (GT > PT > DP), with lower-performing students experiencing greater cognitive load and performance degradation than high-performing students when using the same interfaces. '
-'Typical performance measures of cognitive load have included time to complete tasks, reaction time, correct solutions, memory retrieval time and correctness, time estimation, rate of physical activity and speech, spoken disfluencies, multimodal integration patterns, and other indices'
- 'Dual-task methods are especially relevant and ecologically valid when applied to field and mobile interface design, which chronically involve multitasking and divided attention.'
- think aloud protocol was used to assess whether learners were working on a low level of high-level issue
Oviatt, COulston, Lonsford 2004:When Do We Interact Multimodally? Cognitive Load and Multimodal Communication Patterns
- 'one critical objective of all mobile interface design is the need to manage multitasking, interruption, attentional distraction, fluctuations in the difficulty of natural field tasks and situations, and resulting cognitive overload [8, 18]. This is essential because mobile users need to focus on complex primary field tasks that can vary substantially in difficulty and also involve dual tasking between the field task and secondary tasks involved in controlling an interface. '
- users were using multimodal pen and word interfaces, but findlings are interesting, more multimodal interactions were found when working in a familiar area and when task was more difficult. It is possible that users maintain the working memory resources and distribute tasks to more centres when the task becomes difficult.
Sweller, Merrienboer, Paas 1998:Cognitive Architecture and Instructional Design
- in this paper mechanisms concerning instructional procedures are described in depth: the goal free effect, worked example affect, completion problem effect, split attention effect modality effect, redundancy effect, variability affects,
Papers 3 - When animation benefits from User Control
A. learner control being effective:
Mayer & Chandler, 2001; When Learning Is Just a Click Away: Does Simple User Interaction Foster Deeper Understanding of Multimedia Messages?
- students were presented twice with animation explaining how the storm begins. Some students could control the speed of animation the first time around while the others were controlling the second time around. Having control at the beginning facilitating better learning.
- This supports cognitive load thory: when there is more spare mental load (low extraneous load when animation is not rushed) there is more resources for learning and creating schemas (germane load)
-introducing interactivity and as such did not improve learning.
Hasler et al., 2007; Learner Control, Cognitive Load and Instructional Animation
-'The animations were either system-paced using a continuous animation, learner-paced using discrete segments or learner paced using ‘stop’ and ‘play’ buttons. The two learner-paced groups showed higher test performance with relatively lower cognitive load compared to the two system-paced groups, despite the fact that the ‘stop’ and ‘play’ buttons were rarely used'
-results were obtained only for tasks of higher difficulty
-'Environmental organising and linking principle: Working memory is only limited when dealing with novel information. When dealing with previously organised information from long-term memory, there are no known limits to working memory. Large amounts of information can be brought into working memory from long-term memory and used to determine activity in the environment (Ericsson & Kintsch, 1995).'
-'cognitive load effects disappear when dealing with low element interactivity material (Sweller, 2003, 2004). However, the level of element interactivity is relative to the level of learners’ expertise (Kalyuga, Ayres, Chandler, & Sweller, 2003). For experienced learners, a whole set of interacting elements may be incorporated into a schema and treated as a single element in working memory'
- clt research describes few principles governing cognitive loads: split attention effects , redundancy effect, modality effect, pacing and segmentation
-'learners should be presented information rather than asked to explore a problem space. Accordingly, in the current experiment, learners were presented information on the causes of day and night rather than being asked to search for and explore possible causes.'
-in the second study one group has seen the continues animation twice while the other one has seen a controlled animation twice. Higher learning occured in the controlled condition
-''Learners may closely monitor the information in order to decide whether they need to think about it further by pausing the animation. [...] In contrast, the continuous animation group may have treated the animation as little more than a movie and given it very little thought. '
Schwan & Riempp, 2004:The cognitive benefits of interactive videos: learning to tie nautical knots
- users were presented with either interactive or not interactive movie of how to tie for different sailing knots. Students in the interactive condition took much less time (up to twice shorter) to learn the knots and interaction was much more frequently used when learning the more difficult knots.
-Students in interactive condition did not have more time. 'the effect of requiring more time to watch the more difficult parts of the video repeatedly, in slow motion, or with stops was more or less compensated by skipping the more easy parts of the videos or watching them in speeded mode. But although the viewing times were roughly the same, the viewers of the interactive videos developed a better understanding of the depicted processes than the viewers of the non-interactive videos'
-interactive application was keeping a log of all users interactions: number of stops to practice, number of stops without practising, slow motions, time lapses, number of the direction changes, variability of viewing time across the media.
B. learner control Mixed results
Kriz & Hegarty, 2007:Top-down and bottom-up influences on learning from animations
-in four studies participants were presented with diagrams or animations explaining a mechanical device (European toilet flush) with varying interactive controls and signalling devices. It was found that enhancing animation does not facilitate learning
- this is a very passionately written paper with interesting design (too little details to tell exactly how it workes). Very difficult task selected serves the purpose of escalating the cognitive load needed to analyse an animation way above the threshold. Also the animations might have been just the diagram with annotations (not obvious), so they could not use all the possibilities of animation medium.
Moreno & Valdez, 2005:Cognitive Load and Learning Effects of Having Students Organize Pictures and Words in Multimedia Environments: The Role of Student Interactivity and Feedback
- authors are trying to confirm the theory that 'learning improves when students make, rather than take, meaning'. Participants are learning about the Genesis of storm and either have to learn from cards arranged in a correct order or first ordered them and then learn. Because of a bad application design (intermediate correct/wrong feedback ), users employed trial and error method to organising the cards, and simply had less time learning. In the second study the error was corrected and additionally students were asked to explain why do they think that this is the correct order before they got any feedback on their ordering. He was found at being asked to explain facilitated learning.
- This paper is an interesting history of failure, and proves that reflecting on learning facilitates learning
C learner control frameworks and other:
Moreno & Mayer, 2007; Interactive Multimodal Learning Environments
-the authors describe five design principles: guided activity, reflection, feedback, control, and pretraining.
- Students learn better when allowed to interact with pedagogical agent who helps guide their cognitive processing, when asked to reflect upon correct answers, when getting explanatory rather than corrective feedback, when allowed to control the pace presentation, when they receive focused training
Lowe (2006) - Educational animation: Who should call the shots?
- Author argues that currently designed animations are overwhelming users and the more prescriptive and analytical methods needs to be devised.
- Giving user control might be bad because it introduces an additional task and beginner users might neglect some important details or inappropriately allocate the attention.
- Ideal presentation should emphasise, repeat and slowdown relevant sections without overwhelming users with too much control.
Love 2005 :Changing Perceptions of Animated Diagrams
-very simple and graceful study. Participants are presented with the recording of Newton's cradle, either in normal speed for a half speed. The eye movement patterns and voice descriptions are recorded and analysed.
- From vocabulary analysis it was found that slow motion condition participants were more analytical and explained more, they also looked at the system as a set of items rather than one item. The same effect was found in the eye tracker data.
Domagk, Schwartz & Plass (2010):Interactivity in multimedia learning: An integrated model
- interactivity is not a very well-defined concept and elements that need to be taken under consideration are: behavioural activity, learning environment, cognitive activity, a emotion/motivation, mental model.
Mayer & Chandler, 2001; When Learning Is Just a Click Away: Does Simple User Interaction Foster Deeper Understanding of Multimedia Messages?
- students were presented twice with animation explaining how the storm begins. Some students could control the speed of animation the first time around while the others were controlling the second time around. Having control at the beginning facilitating better learning.
- This supports cognitive load thory: when there is more spare mental load (low extraneous load when animation is not rushed) there is more resources for learning and creating schemas (germane load)
-introducing interactivity and as such did not improve learning.
Hasler et al., 2007; Learner Control, Cognitive Load and Instructional Animation
-'The animations were either system-paced using a continuous animation, learner-paced using discrete segments or learner paced using ‘stop’ and ‘play’ buttons. The two learner-paced groups showed higher test performance with relatively lower cognitive load compared to the two system-paced groups, despite the fact that the ‘stop’ and ‘play’ buttons were rarely used'
-results were obtained only for tasks of higher difficulty
-'Environmental organising and linking principle: Working memory is only limited when dealing with novel information. When dealing with previously organised information from long-term memory, there are no known limits to working memory. Large amounts of information can be brought into working memory from long-term memory and used to determine activity in the environment (Ericsson & Kintsch, 1995).'
-'cognitive load effects disappear when dealing with low element interactivity material (Sweller, 2003, 2004). However, the level of element interactivity is relative to the level of learners’ expertise (Kalyuga, Ayres, Chandler, & Sweller, 2003). For experienced learners, a whole set of interacting elements may be incorporated into a schema and treated as a single element in working memory'
- clt research describes few principles governing cognitive loads: split attention effects , redundancy effect, modality effect, pacing and segmentation
-'learners should be presented information rather than asked to explore a problem space. Accordingly, in the current experiment, learners were presented information on the causes of day and night rather than being asked to search for and explore possible causes.'
-in the second study one group has seen the continues animation twice while the other one has seen a controlled animation twice. Higher learning occured in the controlled condition
-''Learners may closely monitor the information in order to decide whether they need to think about it further by pausing the animation. [...] In contrast, the continuous animation group may have treated the animation as little more than a movie and given it very little thought. '
Schwan & Riempp, 2004:The cognitive benefits of interactive videos: learning to tie nautical knots
- users were presented with either interactive or not interactive movie of how to tie for different sailing knots. Students in the interactive condition took much less time (up to twice shorter) to learn the knots and interaction was much more frequently used when learning the more difficult knots.
-Students in interactive condition did not have more time. 'the effect of requiring more time to watch the more difficult parts of the video repeatedly, in slow motion, or with stops was more or less compensated by skipping the more easy parts of the videos or watching them in speeded mode. But although the viewing times were roughly the same, the viewers of the interactive videos developed a better understanding of the depicted processes than the viewers of the non-interactive videos'
-interactive application was keeping a log of all users interactions: number of stops to practice, number of stops without practising, slow motions, time lapses, number of the direction changes, variability of viewing time across the media.
B. learner control Mixed results
Kriz & Hegarty, 2007:Top-down and bottom-up influences on learning from animations
-in four studies participants were presented with diagrams or animations explaining a mechanical device (European toilet flush) with varying interactive controls and signalling devices. It was found that enhancing animation does not facilitate learning
- this is a very passionately written paper with interesting design (too little details to tell exactly how it workes). Very difficult task selected serves the purpose of escalating the cognitive load needed to analyse an animation way above the threshold. Also the animations might have been just the diagram with annotations (not obvious), so they could not use all the possibilities of animation medium.
Moreno & Valdez, 2005:Cognitive Load and Learning Effects of Having Students Organize Pictures and Words in Multimedia Environments: The Role of Student Interactivity and Feedback
- authors are trying to confirm the theory that 'learning improves when students make, rather than take, meaning'. Participants are learning about the Genesis of storm and either have to learn from cards arranged in a correct order or first ordered them and then learn. Because of a bad application design (intermediate correct/wrong feedback ), users employed trial and error method to organising the cards, and simply had less time learning. In the second study the error was corrected and additionally students were asked to explain why do they think that this is the correct order before they got any feedback on their ordering. He was found at being asked to explain facilitated learning.
- This paper is an interesting history of failure, and proves that reflecting on learning facilitates learning
C learner control frameworks and other:
Moreno & Mayer, 2007; Interactive Multimodal Learning Environments
-the authors describe five design principles: guided activity, reflection, feedback, control, and pretraining.
- Students learn better when allowed to interact with pedagogical agent who helps guide their cognitive processing, when asked to reflect upon correct answers, when getting explanatory rather than corrective feedback, when allowed to control the pace presentation, when they receive focused training
Lowe (2006) - Educational animation: Who should call the shots?
- Author argues that currently designed animations are overwhelming users and the more prescriptive and analytical methods needs to be devised.
- Giving user control might be bad because it introduces an additional task and beginner users might neglect some important details or inappropriately allocate the attention.
- Ideal presentation should emphasise, repeat and slowdown relevant sections without overwhelming users with too much control.
Love 2005 :Changing Perceptions of Animated Diagrams
-very simple and graceful study. Participants are presented with the recording of Newton's cradle, either in normal speed for a half speed. The eye movement patterns and voice descriptions are recorded and analysed.
- From vocabulary analysis it was found that slow motion condition participants were more analytical and explained more, they also looked at the system as a set of items rather than one item. The same effect was found in the eye tracker data.
Domagk, Schwartz & Plass (2010):Interactivity in multimedia learning: An integrated model
- interactivity is not a very well-defined concept and elements that need to be taken under consideration are: behavioural activity, learning environment, cognitive activity, a emotion/motivation, mental model.
Subscribe to:
Posts (Atom)