Speier, Vessey, Valacich 2003 - the effects of interruptions, as complexity, and information presentation on computer supported decision-making performance
–"the interruptions facilitate performance an simple tasks, while inhibiting performance on more complex tasks"
–" special presentation formats were able to mitigate the effects of interruption while symbolic formats were not"
-interruptions have been defined as uncontrollable unpredictable stressors that produce information overload, requiring additional decision-makers effort they require immediate attention and insist on action.
–The difference between distractions and interruptions. Distractions are detected by different sensory channel from the one used for primary primary task and maybe ignored, interactions however use the same sensory trial for both the interruption and the primary task, does decision maker cannot choose to ignore interaction clues. This might need to laws of memory contents and confusion, negatively influencing performance
–distraction conflict theory suggests that distractions facilitate performance on simple tasks and inhibit performance on complex tasks. The narrowing of attention likely results in a decision maker processing fewer information cues, some of which may be relevance to completing tasks successfully
- interruptions and presentation format: a match or cognitive fit occurs when the information emphasised in the particular presentation format (symbolic or spatial) matches that required to most easily complete the task. Symbolic format (tables) are good for presenting discrete set of symbols while spatial format (graphs) is better for depicting relationships among discrete sets of symbols
Schmitz, Heinz, Metrailler,Opwis 2009: cognitive load in e-commerce applications–measurement and effects of user satisfaction
–guidelines for designing usable interfaces recommend reducing short-term memory load. NASA TLX and secondary task (visual monitoring) were used to assess usage of four online bookstores. Users were not allowed to use search, but only to browse the menus. At the end users were asked how much would they prefer search feature then just browsing. While there was no significance on secondary task the high correlation between NASA test and browse/search preference suggests that browse/search preference could be used as a heuristic for assessing how good the browsing environment is.
–Well-designed websites take up more of the cognitive resources for an experienced users, it can be explained by users being able to focus more.
–It's easily lead in this study the secondary task was too difficult and badly placed on the screen
Todd, Benbasat 1994 : the influence of decision aids on choice strategies under conditions of high cognitive load
–According to cost benefit framework the joint objectives of decision-makers are to maximise accuracy and to minimise effort. Often referred may be weighted more heavily than accuracy because feedback in an effort is relatively immediate.
–In many studies effort reduction was far more important than accuracy maximisation, but this might be because effort is the only salient variable. In settings where the decision maker must live with the consequences of the decisions it may be that the decision quality becomes the primary concern.
Hong THing Tam 2003: designing product listing pages on e-commerce websites: and examination of presentation mode and information format
–unclear paper but the really interesting methodology. A mockup website for shopping for groceries contained products of 20 made up brands. In one condition products were shown with the picture in another condition they were shown without a picture. Every time the brand name was shown exactly the same way, and the participants were tested on how well the remembered brand names. It was found that brand names in the pictured condition were remembered better.
Nicholson et all 2005: using distraction conflicts theory to measure the effect of distractions an individual task performance in a wireless mobile environment
- A group of students was given laptops and asked to perform Excel tasks of different difficulty while being in the same room in the cafeteria setting. From the projector different distractions were shown on the wall in front of students eyes: high distraction was a video of a group of students talking about holiday, low distraction was a news TV, no distraction was nothing.
- It was found that performance decreased only enter high distraction condition and only for the high difficulty task.
- It is suggested that mobile wireless work [laptops] is more suitable for simple or well-trained tasks, while more complex tasks should be performed in a more controlled setting, such as an office.
Wednesday, 25 May 2011
Friday, 20 May 2011
Literature 1 - Cognitive load measurement, and how it could influence interface design
Over last few days I was reading a bunch of papers from the Cognitive Load Theory area that could have something to do with my work. Most of them had very interesting methodology. Below are few notes about each:
[currently I am investigating a new line of literature, where business people implement the same techniques as CLT to investigate interfaces which are interruption-safe in working conditions, i.e. Speier, Vessey, Valacich 2003]. I will blog about it next. I also had a meeting With Nadine, where we discussed possible avenues and designs of experiments with shoogle. I will blog about them early next week]
// LITERATURE 1 ///
Chandler, Sweller 1996, Cognitive load while learning to use computer
- 'Hypothesizes that instructional design only takes on a crucial role when there is a high level are in the of interaction between learning elements, resulting in those elements having to be simultaneously held in working memory'
- secondary task used was a program running on a second screen which every eight seconds was attracting attention with the sound and displaying a letter for two seconds. participants had to memorise that letter and in the same time say out loud what the previous letter was
–three learning scenarios were used presenting exactly the same learning information (how to use a software package) in three different types of tutorials. tutorials differed by how much split attention split occurred (extraneous cognitive load) when they were used ( one with words and screenshots which didn't requiring using of computer [low load], one only with commands which were to be followed on a computer screen, and one with text and screenshots which had to be followed and compared with computer [highest split/load]).
- secondary task, , confirmed the order of cognitive load ratings, while skills test confirmed that for complicated tasks ( those with high element interactivity, connected in the long chains of actions) were learned much better with tutorials which did not require a lot of split of attention.
Salomon, '84 ' television is easy and print this tough':
- children were asked about their perceived usefulness of TV and text for learning. They were also asked about realism and possible reasons to fail and succeed in learning with that media. After the children were presented at TV story and equivalent text description.
– It was found that efficacy (usefulness) was in negative correlation with TV and positive with next. Children who rated TV as easy invested less mental effort and hvad worse results. This suggests that there is a connection between preconceptions about interface and its usefulness.
Sweller 2010, element interactivity and intrinsic, extraneous and germane cognitive Load:
- author argues that element interactivity is a defining mechanism not only of intrinsic cognitive load but also underlies extraneous could be developed. ' The same information may impose an intrinsic or an extraneous cognitive load depending on what needs to be learnt'. In other words explanation which is important for a novice can be just slowing down and obscuring learning material for an expert.
- In turn germane cognitive load may not be our separate source of cognitive load, but rather resources used up to handle element interactivity. When the useless interactivity is added the extra use load increases while germane decreases - material seems to be messy, and patterns can be created.
- if Materials are too easy then to remain cognitive load is not activated.
–It might be useful to measure cognitive load during the task rather than after and before
–this possible to determine overall cognitive load, but it may be difficult to determine which cognitive load is of what kind
–goal free approach seems to activate more germane than extraneous load, because learners explored by creating schemas, rather than forcing the information in.
- Introducing interactive elements separately and then joining them produces better learning outcomes than presenting old elements in the same time, in other words it might be useful to train participants in using every element of the interface separately.
–Modality effects can be explained by assigning more resources, while imagination effects increases germane and intrinsic mode only for experts
Chan, Black 06, direct manipulation animation: incorporating the haptic channel in the learning process the suburbs middle school students in science learning a mental model position:
- text, visual and a game context were used to teach children physics problems of different difficulty. ' Findings showed that as the content became increasingly complicated (i.e. increased levels of interactivity between system entities) game like animation proved to be an effective support for students to comprehend the content'.
- This might provide proof that using purpose-built interfaces is suitable only for difficult tasks
Abrams, Davoli 2007: altered vision near the hands:
–participants were asked to perform different vision tasks while their hands were positioned right next to the monitor and when they were positioned at the participants knees.
- In visual search task participants performed better with your hands next to screen only when the task was difficult. During inhibition of return task participants took longer to disengage their attention with their hands next to the monitor. During the attention blink experiment proximity of the screen had negative effect on second task performance right after the first task happened. - attention blink was longer and more severe when object is near our hands.
Wagner, Nusbaum 2004. Probing the mental representation of gesture columns is handwriting spatial?:
- Groups of participants were asked to explain mathematical problems while using or not using gestures. Those using gestures had better learning outcomes.
–He was found that propositional (carrying a meaning) gestures were more useful than simply pointing.
Brunken, Steinbacher, et al 2002 - assessment of cognitive load in multimedia learning using dual task methodology
–when a task is more difficult it requires more cognitive resources, which can be measured by performance on a secondary task.
–"To evaluate variant of interface design drafts for a multimedia system that differ with respect to the format in which information is presented, it is conceivable that this approach could be used to determine the amount of cognitive load induced in the user by each design. On the basis of such an economics of the revelation one could determine for example which interface design variant minimises cognitive load"
[currently I am investigating a new line of literature, where business people implement the same techniques as CLT to investigate interfaces which are interruption-safe in working conditions, i.e. Speier, Vessey, Valacich 2003]. I will blog about it next. I also had a meeting With Nadine, where we discussed possible avenues and designs of experiments with shoogle. I will blog about them early next week]
// LITERATURE 1 ///
Chandler, Sweller 1996, Cognitive load while learning to use computer
- 'Hypothesizes that instructional design only takes on a crucial role when there is a high level are in the of interaction between learning elements, resulting in those elements having to be simultaneously held in working memory'
- secondary task used was a program running on a second screen which every eight seconds was attracting attention with the sound and displaying a letter for two seconds. participants had to memorise that letter and in the same time say out loud what the previous letter was
–three learning scenarios were used presenting exactly the same learning information (how to use a software package) in three different types of tutorials. tutorials differed by how much split attention split occurred (extraneous cognitive load) when they were used ( one with words and screenshots which didn't requiring using of computer [low load], one only with commands which were to be followed on a computer screen, and one with text and screenshots which had to be followed and compared with computer [highest split/load]).
- secondary task, , confirmed the order of cognitive load ratings, while skills test confirmed that for complicated tasks ( those with high element interactivity, connected in the long chains of actions) were learned much better with tutorials which did not require a lot of split of attention.
Salomon, '84 ' television is easy and print this tough':
- children were asked about their perceived usefulness of TV and text for learning. They were also asked about realism and possible reasons to fail and succeed in learning with that media. After the children were presented at TV story and equivalent text description.
– It was found that efficacy (usefulness) was in negative correlation with TV and positive with next. Children who rated TV as easy invested less mental effort and hvad worse results. This suggests that there is a connection between preconceptions about interface and its usefulness.
Sweller 2010, element interactivity and intrinsic, extraneous and germane cognitive Load:
- author argues that element interactivity is a defining mechanism not only of intrinsic cognitive load but also underlies extraneous could be developed. ' The same information may impose an intrinsic or an extraneous cognitive load depending on what needs to be learnt'. In other words explanation which is important for a novice can be just slowing down and obscuring learning material for an expert.
- In turn germane cognitive load may not be our separate source of cognitive load, but rather resources used up to handle element interactivity. When the useless interactivity is added the extra use load increases while germane decreases - material seems to be messy, and patterns can be created.
- if Materials are too easy then to remain cognitive load is not activated.
–It might be useful to measure cognitive load during the task rather than after and before
–this possible to determine overall cognitive load, but it may be difficult to determine which cognitive load is of what kind
–goal free approach seems to activate more germane than extraneous load, because learners explored by creating schemas, rather than forcing the information in.
- Introducing interactive elements separately and then joining them produces better learning outcomes than presenting old elements in the same time, in other words it might be useful to train participants in using every element of the interface separately.
–Modality effects can be explained by assigning more resources, while imagination effects increases germane and intrinsic mode only for experts
Chan, Black 06, direct manipulation animation: incorporating the haptic channel in the learning process the suburbs middle school students in science learning a mental model position:
- text, visual and a game context were used to teach children physics problems of different difficulty. ' Findings showed that as the content became increasingly complicated (i.e. increased levels of interactivity between system entities) game like animation proved to be an effective support for students to comprehend the content'.
- This might provide proof that using purpose-built interfaces is suitable only for difficult tasks
Abrams, Davoli 2007: altered vision near the hands:
–participants were asked to perform different vision tasks while their hands were positioned right next to the monitor and when they were positioned at the participants knees.
- In visual search task participants performed better with your hands next to screen only when the task was difficult. During inhibition of return task participants took longer to disengage their attention with their hands next to the monitor. During the attention blink experiment proximity of the screen had negative effect on second task performance right after the first task happened. - attention blink was longer and more severe when object is near our hands.
Wagner, Nusbaum 2004. Probing the mental representation of gesture columns is handwriting spatial?:
- Groups of participants were asked to explain mathematical problems while using or not using gestures. Those using gestures had better learning outcomes.
–He was found that propositional (carrying a meaning) gestures were more useful than simply pointing.
Brunken, Steinbacher, et al 2002 - assessment of cognitive load in multimedia learning using dual task methodology
–when a task is more difficult it requires more cognitive resources, which can be measured by performance on a secondary task.
–"To evaluate variant of interface design drafts for a multimedia system that differ with respect to the format in which information is presented, it is conceivable that this approach could be used to determine the amount of cognitive load induced in the user by each design. On the basis of such an economics of the revelation one could determine for example which interface design variant minimises cognitive load"
Monday, 16 May 2011
First post: what I was up to so far
This blog is mainly for myself and my supervisor Mike so we can keep track of what i am doing and how cooperation with Nadine works on everyday basis.
So far I was working here for almost 3 weeks from 25th of May with few days off. I've had my workspace setup and all the equipment is here for me to use, UNSW tech people are helpful, over few lunchtimes I've been gathering participants for a friend and it actually turns out that it is not so difficult here as well (unless you require lab and strict experimental conditions, in which case it is very tricky, or so they say).
So far I had few meetings with Nadine and we discussed few possible ways that my work could go into to use her experience and research. It seems that at least 1,2 experiments could be run which can tie our interests together. We recon that during my stay here those experiments will be designed, and I will run them at HW. If results are promising we will write them up with Nadine and try to publish. Right now I found a large bunch of papers which help me to scan the area, and I am making my way trough them. Next meeting is on Thursday, and before this meeting I will write a short note about all of those papers here.
In the meantime I am also changing my iPad app for Liina, a MSc student from London (Nadia's). After few days of on and off work it turned out that she needs too many samples to be displayed, and memory didn't take it well in current design of the app :( This all happened partly because I wanted this app to be more universal, and reusable (common mistake, I guess). That's why today I will start and bash up something quick that will meet all her needs, but not much more. For that purpose I edited even further the sounds that we've been working on with Douglas. They sound quite well by now.
I will keep this blog updated, probably on daily basis. With any questions, just email me.
So far I was working here for almost 3 weeks from 25th of May with few days off. I've had my workspace setup and all the equipment is here for me to use, UNSW tech people are helpful, over few lunchtimes I've been gathering participants for a friend and it actually turns out that it is not so difficult here as well (unless you require lab and strict experimental conditions, in which case it is very tricky, or so they say).
So far I had few meetings with Nadine and we discussed few possible ways that my work could go into to use her experience and research. It seems that at least 1,2 experiments could be run which can tie our interests together. We recon that during my stay here those experiments will be designed, and I will run them at HW. If results are promising we will write them up with Nadine and try to publish. Right now I found a large bunch of papers which help me to scan the area, and I am making my way trough them. Next meeting is on Thursday, and before this meeting I will write a short note about all of those papers here.
In the meantime I am also changing my iPad app for Liina, a MSc student from London (Nadia's). After few days of on and off work it turned out that she needs too many samples to be displayed, and memory didn't take it well in current design of the app :( This all happened partly because I wanted this app to be more universal, and reusable (common mistake, I guess). That's why today I will start and bash up something quick that will meet all her needs, but not much more. For that purpose I edited even further the sounds that we've been working on with Douglas. They sound quite well by now.
I will keep this blog updated, probably on daily basis. With any questions, just email me.
Subscribe to:
Posts (Atom)