Publications

ID Subj S2 Type Title Location Author Year Citation Abstract
52 Nature Amphibians Conference Presentation - Oral Things that go Ribbit in the Night: Amphibian Call Identification Maryland Association for Outdoor and Environmental Education J. Campbell 2011 Campbell, Jeffrey D. (2011). Things that go Ribbit in the Night. Maryland Association for Outdoor and Environmental Education Conference, College Park, MD. Abstract
In many ways, amphibians serve as “canaries in the mine” warning of declining environmental quality. There are less than two dozen species of toads and frogs found in Maryland. Most are easy to identify by their mating calls. The workshop will present the most commonly heard species with species information and recorded calls. Details of two national/regional citizen science programs will be discussed to encourage involvement and sharing of data. Back by Popular Demand
44 Research Plants Conference Poster Volunteer-Driven Restoration within the Maryland Chesapeake Bay National Estuarine Research Reserve 13th Annual Tributary Strategy Teams Meeting P. Delgado and J. Campbell 2009 Delgado, Patricia and Jeffrey D. Campbell (2009). Volunteer-Driven Restoration within the Maryland Chesapeake Bay National Estuarine Research Reserve. 13th Annual Tributary Strategy Teams Meeting, Baltimore, MD. update pending
70 HCI Usability Conference Poster Evaluation of Collaborative GIS Usage with Floor Control Symposium on Computer-Human Interaction for Management of Information Technology (CHIMIT) J. Campbell and M. McGuire 2009 Campbell, Jeffrey D. and Michael McGuire (2009). Evaluation of Collaborative GIS Usage with Floor Control. Symposium on Computer-Human Interaction for Management of Information Technology (CHIMIT), Baltimore, MD. Abstract
Remote real-time shared displays of information in a Geographic Information System (GIS) user interface can provide valuable support to multidisciplinary research teams spread across different locations, emergency management personnel, and teams in the field. In this usability study, small groups of people used standard computer applications - off the shelf GIS and remote control software - to collaboratively perform spatial search and analytical tasks. This system architecture implements a floor control policy specifying only one user at a time controls the input but all users immediately see the results.
The objective of this research is to improve the design of remote real-time shared GIS software based on empirical data focused on user-software interaction in a realistic scenario. The focus is on human-computer interaction. The participants in the study evaluated the system in terms of ease of use and related factors.
Key results include favorable responses about using shared GIS, overwhelming interest in using it again, and mitigation of expected floor control and navigation difficulties. Further study with more than two users and with varied tasks is recommended.
Conference Program
68 Environmental Informatics Water Qualilty Conference Poster Comparison of Volunteer and Automated Water Quality Data Suggesting Procedural Improvements Maryland Water Monitoring Council Annual Conference J. Campbell, L. Hollister, and R. Smith 2009 Campbell, Jeffrey D., Lindsay F. Hollister and Robert Smith (2009). Comparison of Volunteer and Automated Water Quality Data Suggesting Procedural Improvements. Maryland Water Monitoring Council Annual Conference, Linthicum, MD. Described in the Environmental Information, QA/QC
69 Environmental Informatics Water Qualilty Conference Poster Using Our Multi-Source Water Quality Database for QA/QC Analysis EPA Region 3 Volunteer Water Monitoring Conference J. Campbell, L. Hollister, and R. Smith 2009 Campbell, Jeffrey D., Lindsay F. Hollister and Robert Smith (2009). Using Our Multi-Source Water Quality Database for QA/QC Analysis. EPA Region 3 Volunteer Water Monitoring Conference, Carlisle, PA. Described in the Environmental Information, QA/QC
43 Research Wetlands Conference Poster Understanding the Fate of the Jug Bay Fresh Water Tidal Wetlands in Light of Sea Level Rise: A Conceptual Model Atlantic Estuarine Research Society 2009 Spring Meeting P. Delgado, C. Swarth, E. Freible, L. Harris and J. Campbell 2009 Delgado, Patricia, Christopher Swarth, Elaine Freible, Lora Harris and Jeffrey D. Campbell (2009). Understanding the Fate of the Jug Bay Fresh Water Tidal Wetlands in Light of Sea Level Rise: A Conceptual Model. Atlantic Estuarine Research Society 2009 Spring Meeting, Ocean City, MD. update pending
66 Environmental Informatics Warehouse Invited Talk Stalking Wild Data: Environmental Informatics at Jug Bay Chesapeake Biological Laboratory, University of Maryland J. Campbell 2009 Campbell, Jeffrey D. (2009). Stalking Wild Data: Environmental Informatics at Jug Bay. Brown Bag Seminar, July 16, 2009, Chesapeake Biological Labratory, University of Maryland Center for Environmental Science, Solomons Island, MD. update pending
65 Environmental Informatics GIS Invited Talk Collaborative Geographic Information Systems Analysis" International Symposium on Collaborative Technologies and Systems J. Campbell 2009 Campbell, Jeffrey D. (2009). Collaborative Geographic Information Systems Analysis. International Symposium on Collaborative Technologies and Systems, Baltimore, MD. update pending
53 Nature Amphibians Conference Presentation - Oral Things that go Ribbit in the Night: Amphibian Call Identification Maryland Association for Outdoor and Environmental Education J. Campbell 2010 Campbell, Jeffrey D. (2009). Things that go Ribbit in the Night. Maryland Association for Outdoor and Environmental Education Annual Conference, Rocky Gap, MD, February 4-7, 2010. update pending
35 Other Volunteer Newsletter Volunteer Research Symposium PDF Volunteer Monitor C. Swarth 2008 Swarth, Christopher (2008). Volunteer Research Symposium. Volunteer Monitor. Vol 19, No 1, pp 6 - 7. A description of the Jug Bay Wetlands Research Symposium
64 Environmental Informatics Warehouse Conference Presentation - Oral Stalking Wild Data: Environmental Informatics at Jug Bay Jug Bay Wetlands Research Symposium J. Campbell 2008 Campbell, Jeffrey D. (2008). Stalking Wild Data: Environmental Informatics at Jug Bay. Jug Bay Wetlands Research Symposium, Annapolis, MD, March 29, 2008. update pending
63 Research Plants Conference Poster Wild Rice 2007: Lessons Learned Jug Bay Wetlands Research Symposium J. Campbell, Y.Tasumi and E. Freible 2008 Campbell, Jeffrey D., Yuka Tasumi and Elaine Freible (2008). Wild Rice 2007: Lessons Learned. Jug Bay Wetlands Research Symposium, Annapolis, MD, March 29, 2008. update pending
57 Environmental Informatics Warehouse Newsletter Stalking Wild Data: Environmental Informatics at Jug Bay Marsh Notes J. Campbell 2008 Campbell, Jeffrey D. (2008). Stalking Wild Data: Environmental Informatics at Jug Bay. Marsh Notes. Vol 22, No 1, pp 10 - 11. Quarterly newsletter, Jug Bay Wetlands Sanctuary
54 Environmental Informatics Data Management Invited Presentation Environmental Data Management Chesapeake Bay National Estuarine ResearchReserve - Site Managers quarterly meeting J. Campbell 2008 Campbell, Jeffrey D. (2008). Environmental Data Management. Chesapeake Bay National Estuarine ResearchReserve - Maryland Site Managers quarterly meeting, August 25, 2008 update pending
45 Research Amphibians Conference Poster Monitoring Protocol for Larger Vernal Pools Vernal Pool Workshop, Maryland Water Monitoring Council J. Campbell, L.Hollister and E. Freible 2008 Campbell, Jeffrey D., Lindsay F. Hollister and Elaine Freible (2008). Monitoring Protocol for Larger Vernal Pools. Vernal Pool Workshop, Maryland Water Monitoring Council,, Lusby, MD. More information in the Maryland Nature, Vernal Pools page links
37 Environmental Informatics Database Conference Presentation - Oral Database Development: Best Practices Case Study Maryland Water Monitoring Council Annual Conference J. Campbell 2008 Campbell, Jeffrey D. (2008). Database Development: Best Practices Case Study. Maryland Water Monitoring Council Annual Conference, Linthicum, MD. Annotated slide show and related information on the Information Technology, Fish Survey Database page
Abstract
This presentation will use stream monitoring case studies to illustrate best practices in database design. Starting from the underlying system requirements, the quality of various design alternatives will be evaluated. This evaluation will include both technical aspects and those that will improve ease of use by the people ultimately using the system. Key technical aspects include reducing redundant information to minimize potential inconsistencies, ensuring that related data is easily retrievable, and reducing the risk of unintentionally losing data. Ease-of-use factors include reducing data entry effort, simplifying data entry validation, and enhancing productivity in retrieving data with queries. The best practices are not specific to the database software (e.g. Microsoft Access, MySQL, etc) being used.
Attendees with some prior database experience will be best able to apply the concepts illustrated to their own data. Although the focus is on developing database designs, the concepts will also help attendees create spreadsheets that can be imported into a database more easily at a future time.
36 Environmental Informatics GIS Conference Poster Evaluation of Collaborative GIS Usage American Geophysical Union Fall Meeting J. Campbell and M.McGuire 2008 Campbell, Jeffrey D. and Michael McGuire (2008). Evaluation of Collaborative GIS Usage. American Geophysical Union Fall Meeting, San Francisco, CA. update pending
41 HCI Collaborative Editing Conference Poster submitted by a student How People Work Together Creating Documents Using Real Time Collaborative Text Editors APA Div 21, Div 19 and HFES Potomac Chapter Annual Symposium on Applied Experimental Research S.Puckett and J. Campbell 2007 Puckett, Seth and Jeffrey D. Campbell (2007). How People Work Together Creating Documents Using Real Time Collaborative Text Editors. American Psychological Association Division 21, Division 19 and Human Factors and Ergonomic Society Potomac Chapter Annual Symposium on Applied Experimental Research, Fairfax, VA. Real-time collaborative text editors allow people at different locations to view and modify a document at the same time. Each person can immediately see the changes made by anyone. Due to computer network delays, the copies of the document could fail to be identical. Most of the recently published research has focused on enhancements to techniques to resolve this technical problem. However, little research has been done to see how people actually use the software. Surprisingly, the frequency with which the technical problem occurs has not been published.
A study was conducted with twelve pairs of people using a collaborative editing system to write a document for about a half hour each. The objective was to better understand how the people used the software. Analysis of detailed keystroke logs supplements the observations and usability survey results.
We found that the subjects tended to correct and revise the text they had written (80-90% of changes) instead of what was written by the other person. A second result is that the technical problem is likely to be significant with less than two-tenths of one percent of the keystrokes. This suggests that social solutions may enhance the usability of the system compared to the current arbitrary algorithms that are not sensitive to the work and social context.
Results of this analysis can improve the user interface for collaborative editing software by more closely matching it to the way the software is actually used.

This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
40 Human-Computer Intereaction Location-aware computing Conference Poster submitted by a student Usability Evaluation of Computer Support for Fine-Grained Location-Finding APA Div 21, Div 19 and HFES Potomac Chapter Annual Symposium on Applied Experimental Research M. J. Morelli and J. Campbell 2007 Morelli, Matthew J. and Jeffrey D. Campbell (2007). Usability Evaluation of Computer Support for Fine-Grained Location-Finding. American Psychological Association Division 21, Division 19 and Human Factors and Ergonomic Society Potomac Chapter Annual Symposium on Applied Experimental Research, Fairfax, VA. This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
58 Nature Fish FieldGuide Fish of Jug Bay Patuxent River, Maryland: A Pictorial Key Jug Bay Wetlands Sanctuary K. Molines and J. Campbell 2007 Molines, Karyn and Jeffrey D. Campbell (2007). Fish of Jug Bay Patuxent River, Maryland: A Pictorial Key. Lothian, MD, Jug Bay Wetlands Sanctuary, 18 pages. PDF file and additional description in Maryland Nature, Freshwater Fish
62 Nature Amphibian Newsletter Spring Peepers Still Number One ECOnnection, (Howard County Department of Recreation and Parks Environmental Newsletter) J. Campbell 2007 Campbell, Jeffrey D. (2007). Spring Peepers Still Number One. ECOnnection, (Howard County Department of Recreation and Parks 2007 Environmental Newsletter), page 3. update pending
38 Environmental Informatics Data Mining Conference Poster Data Mining for Ecological Field Research: Lessons Learned from Amphibian and Reptile Activity Analysis NSF Symposium on Next Generation of Data Mining and Cyber-Enabled Discovery for Innovation J. Campbell, K. Molines and C. Swarth 2007 Campbell, Jeffrey D., Karyn Molines and Christopher Swarth (2007). Data Mining for Ecological Field Research: Lessons Learned from Amphibian and Reptile Activity Analysis. NSF Symposium on Next Generation of Data Mining and Cyber-Enabled Discovery for Innovation, Baltimore, MD. Abstract
While there are many similarities to traditional data mining, analysis of ecological field research data pre-sents some uncommon data and algorithmic require-ments. Data challenges reflect the broad range of data that is relevant to analysis in this domain. Much of that data is obtained from physical measurements that are subject to mechanical, electrical or human error factors. For some types of data, for example meteorological ob-servations, multiple sources are publicly available. One challenge is to properly distinguish errors from micro-climatic variations which are important to ecologists. Varied units of measure and inconsistent time policies are also commonly found in environmental data.
Requirements for data mining algorithms for animal activity analysis include computing differential and inte-gral values, handling highly interrelated environmental factors, and strong support of time series analysis. This includes both variable types of lag times and more types of cyclical effects than usually found in business data.
47 Human-Computer Intereaction Instant Messaging Conference Poster submitted by a student Errors in Instant Messaging UMBC Undergraduate Summer Research Fest K. Solomon and J. Campbell 2006 Solomon, Katherine and Jeffrey D. Campbell (2006). Errors in Instant Messaging. UMBC Undergraduate Summer Research Fest, Baltimore, MD. Abstract
Instant Messaging (IM) is an informal tool used both in business and for personal use which is growing in popularity. By measuring the sensitivity to typographical and grammatical errors, new User Interface features or new IM functionality can be developed. This study examines participants’ ability to observe, correct and perceive errors during and after IM sessions. For this study, 45 pairs of students performed one of two tasks. In one task, the students jointly prioritized a list of items needed for survival in an emergency situation. In the other task, they assembled blocks while following the instructions provided by the other person. Their sensitivity to errors was measured both by their perception of errors made, and by their performance in identifying the errors made during their IM task. The students’ rate of uncorrected errors was significantly higher for IM than in a standardized typing test. This is consistent with the informal nature of IM. There was no significant difference in the rate of uncorrected errors in either task. However there was a difference in the perception of the amount of errors and their impact based on which task was being performed. This relationship between task and error sensitivity was not previously detected. This discovery leads to future research about the aspects of the task that help determine error sensitivity. With that information, a more intelligent interface for IM can be developed accommodating the user and the task.

This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
50 Human-Computer Intereaction Collaborative Editing Conference Poster submitted by a student Keystroke Transpositions in Real-Time Collaborative Text Editors UMBC Undergraduate Summer Research Fest S.Puckett and J. Campbell 2006 Puckett, Seth and Jeffrey D. Campbell (2006). Keystroke Transpositions in Real-Time Collaborative Text Editors. UMBC Undergraduate Summer Research Fest, Baltimore, MD. This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
46 Human-Computer Intereaction Location-aware computing Conference Poster submitted by a student Usability Evaluation of Computer-Supported Indoor Location-Finding UMBC Undergraduate Summer Research Fest M. Morelli and J. Campbell 2006 Morelli, Matthew J. and Jeffrey D. Campbell (2006). Usability Evaluation of Computer-Supported Indoor Location-Finding. UMBC Undergraduate Summer Research Fest, Baltimore, MD. This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
56 Environmental Informatics Usability Evaluation Invited Presentation Usability Evaluation Southern California Earthquake Center, University of Southern California J. Campbell 2006 Campbell, Jeffrey D. (2006). Usability Evaluation. Southern California Earthquake Center, University of Southern California, Los Angeles, CA.
39 Human-Computer Intereaction Instant Messaging Conference Poster submitted by a student Perception of Errors in Instant Messaging APA Div 21, Div 19 and HFES Potomac Chapter Annual Symposium on Applied Experimental Research J. Campbell and C. Naclerio 2006 Campbell, Jeffrey D. and Christine Naclerio (2006). Perception of Errors in Instant Messaging. American Psychological Association Division 21, Division 19 and Human Factors and Ergonomic Society Potomac Chapter Annual Symposium on Applied Experimental Research, Fairfax, VA. This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
32 Human-Computer Intereaction Instant Messaging Conference Poster Instant Messaging Comprehension with Non-Keyboard Composition DOI ACM CHI 2006 Extended Abstracts on Human Factors in Computing Systems J. Campbell and C. Naclerio 2006 Campbell, Jeffrey D. and Christine Naclerio (2006). Instant Messaging Comprehension with Non-Keyboard Composition. ACM CHI 2006 Extended Abstracts on Human Factors in Computing Systems, pp 628 - 633, ACM Press. Abstract
A comparison of the accuracy of comprehension of Instant Messaging (IM) messages composed using a keyboard, handwriting recognition and speech recognition was performed. The participants in the study were shown text messages either containing no errors or having errors caused by typing, handwriting or speech recognition. They demonstrated their understanding of the messages by correcting the errors as they retyped the messages. The time to correct typographical errors was not significantly different from the time to type messages without errors. Speech and handwriting recognition errors in the text took significantly longer to correct than typographical errors, and the results were significantly less accurate. Interestingly, participants expressed slightly higher confidence in their understanding of messages containing speech recognition errors than handwriting, but the accuracy results from those two categories were not significantly different.
30 Human-Computer Intereaction Diagram creation Journal Coordination for multi-person visual program development DOI Journal of Visual Languages and Computing J. Campbell 2006 Campbell, Jeffrey D. (2006). Coordination for Multi-Person Visual Program Development. Journal of Visual Languages and Computing 17(1): 46 - 77 Abstract
Typically, visual programming has been limited to only one person developing one program at a time. This article describes a technique for helping multiple people coordinate working together on the same diagram at the same time. This approach identifies transactions based on domain syntax. These transactions are used to notify people when someone else changes the diagram in a way that is likely to impact other people's work. In particular, the system assigns ownership of each syntactically incorrect element to the person who last acted upon that element. This ownership can be transferred between people. The potential problem of incomplete transactions when work extends beyond a single session is resolved by restarting transactions when work resumes. This syntax-based approach is particularly appropriate for visual languages. Various domain constraints are described as alternatives or supplements to the syntactic criteria. The technique was validated with data from 20 groups of three people using CoDiagram, a proof of concept system.
33 Human-Computer Intereaction Data Analysis Workshop Paper Log and Qualitative Analysis of Collaborative Systems" Usage Analysis: Combining Logging and Qualitative Methods Workshop at ACM CHI 2005 J. Campbell 2005 Campbell, Jeffrey D. (2005). Log and Qualitative Analysis of Collaborative Systems. "Usage Analysis: Combining Logging and Qualitative Methods Workshop" at ACM CHI 2005, Portland, OR. This paper describes the author’s interest and experience with collecting and analyzing data from collaborative sys-tems such as instant messaging and real-time shared editors for both text documents and diagrams.
24 Human-Computer Intereaction Instant Messaging Conference Presentation of Paper An Evaluation of Immediate Instant Messaging HCI International Conference J. Campbell, Z. Zhang and R. D. Nolker 2005 Campbell, Jeffrey D., Zhenxue Zhang and Robert D. Nolker (2005). An Evaluation of Immediate Instant Messaging. HCI International, Las Vegas, NV. Abstract
An evaluation of the usability of instant messaging (IM) software with two different interfaces was conducted. One was a standard interface that displayed messages when they were sent. The other interface displayed each character as it was typed by the other person. Participants in the study performed one of two tasks while communicating only via IM. User satisfaction surveys indicated that the participants found the immediate interface to be less pleasant. Participant comments identified positive and negative aspects of the new interface. The effect of the interface on task performance time is small. Message composition statistics show a decrease in simultaneous composition of messages by both participants with a small reduction in discarded text.
23 Human-Computer Intereaction Instant Messaging Conference Poster Does Spelling Matter in Instant Messaging? Answers from Measuring Error Correction Frequency DOI CHI '05 extended abstracts on Human factors in computing systems J. Campbell 2005 Campbell, Jeffrey D. (2005). Does Spelling Matter in Instant Messaging? Answers from Measuring Error Correction Frequency. CHI '05 extended abstracts on Human factors in computing systems, Portland, OR, pp 1244 - 1247, ACM Press. Abstract
Although Instant Messaging (IM) is widely perceived as a communication mode with relaxed spelling and grammar rules, three measures show that people still perceive errors and often correct them. In two usability studies with different tasks, participants on average corrected 89 and 95 percent of their typographical and grammatical errors that they identified. However, their self-correction rates were substantially lower and more variable when compared to all errors found by independent proofreaders. The results from the second measurement showed that participants were aware of their own errors. They judged their own error rate to be significantly higher than the other person's even when that was not true. Using the third measure, correction rates in IM were found to be lower than "normal" typing. The self-sensitivity to errors prompts further investigation of techniques for error detection and correction appropriate for IM's rapid pace and relaxed linguistic register.
49 Human-Computer Intereaction UI Conference Poster submitted by a student Error Recognition from Multiple Input Sources UMBC Undergraduate Summer Research Fest C. Naclerio and J. Campbell 2004 Naclerio, Christine and Jeffrey D. Campbell (2004). Error Recognition from Multiple Input Sources. UMBC Undergraduate Summer Research Fest, Baltimore, MD. This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
21 Human-Computer Intereaction Diagram creation Workshop Experience with Sharing Diagrams Making Application Sharing Easy: Architectural Issues for Collaboration Transparency Workshop at ACM CSCW 2004 J. Campbell 2004 Campbell, Jeffrey D. (2004). Experience with Sharing Diagrams. "Making Application Sharing Easy: Architectural Issues for Collaboration Transparency Workshop" at ACM Computer Supported Cooperative Work Conference (CSCW 2004), Chicago, IL. update pending
20 Human-Computer Intereaction Instant Messaging Conference Presentation of Paper Instant Messaging: A Framework for Reading Between the Lines DOI Proceeding CSCW '04 Proceedings of the 2004 ACM Conference on Computer Supported Cooperative Work J. Campbell 2004 Campbell, Jeffrey D. (2004). Instant Messaging: A Framework for Reading Between the Lines. Proceeding
CSCW '04 Proceedings of the 2004 ACM Conference on Computer Supported Cooperative Work, Chicago, IL, pp 519 - 522,ACM Press.
Abstract
A framework is described for analyzing keystroke level data from instant messages (IM). This is unlike other analyses of IM which employ server-based logs of messages. This framework can be used to identify metrics for evaluating the usability of IM during message composition. The current objective is evaluating awareness features. The model also identifies quantifiable factors that can be computed automatically during IM usage that could allow the system to adapt to different styles of IM usage. Data from a representative usability evaluation scenario is utilized to illustrate some results of using this framework. Computational aspects of the framework have been implemented in GLogger.
19 Human-Computer Intereaction Security Conference Presentation of Paper Enhanced Access Control for Collaborative Documents Secure Knowledge Management Workshop J. Campbell 2004 Campbell, Jeffrey D. (2004). Enhanced Access Control for Collaborative Documents. Secure Knowledge Management Workshop, Amherst, NY, pp 187 - 191. Abstract
A coordination technique [1] from the area of computer supported cooperative work can be enhanced to provide a rich access control mechanism. While the technique was originally developed to allow a system to help multiple people coordinate working on the same document at the same time, the security application would not necessarily involve simultaneous work. The initial implementation was with shared diagrams but the underlying concepts are also applicable to text. The technique uses domain structure or constraints from the application area to coordinate access to diagram shapes . The original purpose was to avoid inconsistent concurrent updates, but the enhancement described here uses the same concepts to accomplish security objectives. In particular, it can control access to meaningful sections of a diagram instead of just using physical proximity. The technique can also provide longer duration access information which can assist in identifying people with interests in a particular area. This can facilitate cooperation or detect unexpected users
17 Human-Computer Intereaction Data Analysis PeerReviewDemo GLogger: A Tool for Collaborative Usability Study Analysis and Visualization Demonstration at CSCW 2004 J. Campbell 2004 Campbell, Jeffrey D. (2004). GLogger: A Tool for Collaborative Usability Study Analysis and Visualization. Demonstration at ACM Computer Supported Cooperative Work Conference (CSCW 2004), Chicago, IL. Abstract
GLogger was used to perform the analysis for the “Instant Mes-sages: A Framework for Reading Between the Lines” CSCW 2004 Note. GLogger supports evaluations of multi-user software. It collects, manages, analyzes and provides visualizations of log and multimedia data from synchronous or asynchronous users. The log files typically contain event data ranging from key-strokes/mouse clicks to application specific events to task activ-ity. The analysis component is particularly adept at identifying sequences or blocks of records that meet particular criteria. GLogger has also been used with collaborative editing data. GLogger supports manual coding, such as categorizing message types, and can perform calculations and analysis based on that coding. Multiple, linked, filterable visualizations of the data are provided including hierarchical, multimedia, time line, time chart, document and data table.
12 Human-Computer Intereaction Diagram creation Journal Interaction in Collaborative Computer Supported Diagram Development DOI Computers in Human Behavior J. Campbell 2004 Campbell, Jeffrey D. (2004). Interaction in Collaborative Computer Supported Diagram Development. Computers in Human Behavior 20(2): 289 - 310. Abstract
A number of computer programs have been developed to allow multiple people to work on the same document at the same time from different computers possibly in different locations. These allow groups to create and modify text and diagrams. A problem for people using these programs is that the physical constraints of working together are removed. For example, there is no pen to share, and everyone can access any part of the diagram without bumping into another person. Without being able to see each other, two people might try to make changes to the same word or shape in a diagram. While this might be good cooperative work, there is a risk that they have different intentions, which results in interfering actions. CoDiagram, a program to allow multi-person drawing of diagrams, was used to measure the interactions between people working together to create a diagram to solve a design problem. Understanding this interaction is an important aspect of designing collaborative software that will be easy to use and that people will like. It was found that the people readily adapted to using the multi-person diagram software. About 65% of the shapes in the diagram were manipulated by more than one of the people in a group, showing widespread sharing of the work. While it was found that interference is relatively rare, the people rated the impact of the interference as significant.
13 Human-Computer Intereaction Instant Messaging Conference Presentation of Paper Instant Messaging: Between the Messages DOI IEEE Systems, Man and Cybernetics Conference J. Campbell, E. Stanziola and J. Feng 2003 Campbell, Jeffrey D., Enrique Stanziola and Jinjuan Feng (2003). Instant Messaging: Between the Messages. IEEE Systems, Man and Cybernetics Conference, Volume 3, pp 2193 - 2198, Washington, DC, IEEE Press. Abstract
Many studies have examined Instant Messaging server logs containing messages that were sent. Such studies cannot address questions about client activity during message composition because the necessary information is not available at the server. This study examines keystroke level data to understand what is happening between the sending of messages. Sixteen pairs of participants used Instant Messaging to solve a problem. Even though this task was designed to fully involve the participants' attention, only 43% of the time was spent actually typing messages. An average of 12% of that composition time was spent revising text that had already been written, often making substantial changes. These results demonstrate an opportunity for improvement of user interface features to better inform the user about the other person's activity in order to improve the efficiency of message composition.
10 Human-Computer Intereaction An Conference Presentation of Paper Data Analysis and Visualization for Usability Evaluation for Collaborative Systems HCI International Conference
J. Campbell, E. Stanziola and A. Sears 2003 Campbell, Jeffrey D., Enrique Stanziola and Andrew Sears (2003). Data Analysis and Visualization for Usability Evaluation for Collaborative Systems. HCI International, Volume 2, pp 869 - 873, Crete, Greece, Lawrence Erlbaum Associates. Abstract
GLogger is a data collection, management and analysis tool specifically designed to meet the requirements for usability evaluations of multi-user collaborative applications. It has been used to collect and analyze data from the development of diagrams, collaborative text editing and instant messaging. It provides support for rich meta-data and synchronization needed for multi-user logging. GLogger is distinguished by its flexibility and multiple visualizations of the data. Data views include hierarchical, multimedia, time line, time chart, document and data table.
48 Human-Computer Intereaction Diagram creation Conference Poster submitted by a student Sharing in a Collaborative Diagramming Environment UMBC Undergraduate Summer Research Fest K. Buffardi and J. Campbell 2003 Buffardi, Kevin and Jeffrey D. Campbell (2003). Sharing in a Collaborative Diagramming Environment. UMBC Undergraduate Summer Research Fest, Baltimore, MD. Abstract
Diagrams are graphical representations of ideas or designs. When people are working simultaneously on one diagram, physical restrictions can easily hinder the productivity of the group by inhibiting access to the drawing. However, software removes the limitations of working on a common, tangible medium and allows elements, or “shapes,” of the diagram to be collaboratively edited. While software may increase efficiency, it also introduces different behaviors in a new working environment. These behaviors must be understood in order to design software that best supports a group of users.
Various measurements were developed to quantify user interaction including: number of unique users who edit a single shape; distribution of shape creation among users; and patterns of shape ownership. Nine groups, each composed of three graduate students, solved a software engineering problem using CoDiagram software. CoDiagram allows users to edit the same diagram from different computers. Whenever one user changes the diagram, the change is immediately updated on all the users’ computer screens. Although none of the participants had previous collaborative diagramming experience, each group was successful in designing a diagram.
Overall, three quarters of the diagram shapes were used by more than one person, demonstrating that significant interaction occurs. Such interaction encourages parallel work, but it also introduces new issues that must be addressed in designing collaborative diagram systems. For example, conflicts can arise if one person uses a shape in a manner inconsistent with another person’s intentions. The system interface needs to replace the physical cues that ordinarily inhibit such interference. A comprehensive understanding of shape usage and sharing can contribute to an effective system design that supports cooperation while minimizing negative consequences.

This material is based upon work supported by the National Science Foundation under Grant No. EIA-0244131. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF.
48 Human-Computer Intereaction Instant Messaging Conference Poster submitted by a student Instant Messaging: Error Tolerance and Perception UMBC Undergraduate Research and Creative Achievement Day A. Adedeji and J. Campbell 2003 Adedeji, Abimbola A and Jeffrey D. Campbell (2003). Instant Messaging: Error Tolerance and Perception. UMBC Undergraduate Research and Creative Achievement Day, Baltimore, MD. Abstract
This study examined the perception of errors during Instant Messaging (IM) between pairs of people engaged in a problem-solving task. Thirty-two participants completed a survey including questions about spelling, grammar and vocabulary errors before they proofread a transcript of the messages. These measures of the participant’s perception of errors were compared to an expert’s analysis of the transcripts. A statistically significant result is that a person is much more likely to perceive errors that they themselves made than errors made by another person. This has important implications for the design of the user interface for IM. In the future, IM could include input via speech or hand writing recognition that have much higher error rates than typing. With such alternative inputs, it will be necessary to include features that correct, or at least identify, errors before they are sent so that the user would not be overly self-conscious about their error rate.
22 Human-Computer Intereaction Diagram creation Conference Presentation of Paper Characteristics of Group Development of Diagrams DOI Evaluating Collaborative Enterprises Workshop at Eleventh IEEE WET-ICE
J. Campbell 2002 Campbell, Jeffrey D. (2002). Characteristics of Group Development of Diagrams. Evaluating Collaborative Enterprises Workshop at Eleventh IEEE International Workshops on Enabling Technologies: Infrastructure for Collaborative Enterprises (WET-ICE 2002), pp 29 - 34, Pittsburgh, PA, IEEE Press. Abstract
One contributing factor to the lack of wide spread usage of software for collaborative development of diagrams is the high degree of variability in the requirements for such software. The requirements are categorized here in terms of diagram characteristics and diagram creation strategies. Various measures are proposed to characterize a potential collaborative diagramming need and to compare it to potential software systems. The categories and measures are also appropriate for evaluating the match between existing software and specific requirements. They can serve to guide future collaborative diagram software development.
9 Human-Computer Intereaction Diagram creation Conference Presentation of Paper Multi-User Collaborative Visual Program Development DOI IEEE Symposia on Human Centric Computing Languages and Environments J. Campbell 2002 Campbell, Jeffrey D. (2002). Multi-User Collaborative Visual Program Development. Proceedings of the IEEE 2002 Symposia on Human Centric Computing Languages and Environments (HCC'02), pp 122 - 130, Arlington, VA. Abstract
Visual programming has traditionally supported only one person working on a program at a time. Researchers in the field of computer supported cooperative work have developed text and graphics editors that allow multiple people to work on the same document at the same time from multiple computers viewing the same, shared document. An unresolved problem is how to avoid interference between the users. A technique based on domain syntax is presented here. It is shown to be effective in preventing such interference during collaborative development of Entity-Relationship diagrams using CoDiagram, a proof of concept system. Twenty groups of thee people developed diagrams and rated the system as usable.
34 OTHER Other Newsletter CHI Students ACM SIG CHI Bulletin Supplement to Interactions E.Orrick 2001 Orrick, Erika (2001). CHI Students. ACM SIG CHI Bulletin Supplement to Interactions July-August 2001: 12. Notes:
Describes CHI 2001 student volunteer program that I co-chaired with Carolyn Gale
14 Human-Computer Intereaction Collaborative Editing Workshop Third Annual Collaborative Editing Workshop DOI ACM SIGGROUP Conference on Supporting Group Work J. Campbell 2001 Campbell, Jeffrey D. (2001). Third Annual Collaborative Editing Workshop. GROUP '01 Proceedings of the 2001 International ACM SIGGROUP Conference on Supporting Group Work, pp 3 -4, Boulder CO, ACM Press. Abstract
Collaborative editing systems support groups of people editing a document together over the computer network. People may work simultaneously on the same document, simultaneously on different copies of the document, or at different times on the original or copied document. The document types include text, diagrams, more complicated graphic objects, images, CAD drawings, multimedia, etc.. Collaborative editing research addresses many fundamental and challenging issues facing the designers of real-time groupware systems in general.
11 Human-Computer Intereaction Collaborative Editing Conference Presentation of Paper Evaluation of Interference During Collaborative Document Development DOI Evaluating Collaborative Enterprises Workshop at Tenth IEEE WET-ICE 2001 J. Campbell 2001 Campbell, Jeffrey D. (2001). Evaluation of Interference During Collaborative Document Development. Evaluating Collaborative Enterprises Workshop at Tenth IEEE International Workshops on Enabling Technologies: Infrastructure for Collaborative Enterprises (WET-ICE 2001), pp 108-113, Cambridge, MA, IEEE Press. Abstract
A problem with collaborative development of documents is that people can interfere with each other?s work. If unresolved, this interference can result in an incorrect document. A evaluation scenario using the development of a database design is proposed as a good context for evaluation of collaborative systems for the development of both text and diagram documents. Several metrics and measures are described for quantifying the occurrence of interference. In particular, five survey questions were added to a standard usability instrument to address interference and protection.
8 Human-Computer Intereaction Collaborative Editing Workshop Collaborative Editing Data Schema Third International Workshop on Collaborative Editing at ACM Group 2001 J. Campbell 2001 Campbell, Jeffrey D. (2001). Collaborative Editing Data Schema. "Third International Workshop on Collaborative Editing" at ACM Group 2001, Boulder, CO. update pending
16 Information Science Document Processing EncArticle Document Processing Encyclopedia of Library and Information Science M. Spring and J. Campbell 2000 Spring, Michael B. and Jeffrey D. Campbell (2000). Document Processing. Encyclopedia of Library and Information Science. New York, Marcel Dekker. 67, Supplement 30, pp 134 - 159. update pending
18 Human-Computer Intereaction Diagram creation Dissertation Consistency Maintenance for Real-Time Collaborative Diagram Development link Information Science Department J. Campbell 2000 Campbell, Jeffrey D. (2000). Consistency Maintenance for Real-Time Collaborative Diagram Development. Department of Information Science, University of Pittsburgh, Pittsburgh, PA, 164 pages, ISBN:9780599799103 Abstract
The current study examines the development of diagrams by groups of people. Diagrams are graphical representations of a design, concept or object. Entity Relationship diagrams are created during a usability study of CoDiagram, a system supporting multiple people working simultaneously on the diagram. Computer support for diagramming allows groups of people to work on the diagram at the same time, improving productivity. However, multiple users working together can interfere with each other's work. During the usability study, it was confirmed that people work with shapes in the diagram that were created by other people, demonstrating the potential for interference. While it was found that interference is relatively rare, the people rated the impact of the interference as significant. At a minimum, the interference results in lost time as they resolve the conflict. If undetected, the interference can cause inconsistencies or errors that can make the diagram practically worthless. The conclusion from this initial phase of research is that a concurrency control mechanism is needed to maintain integrity for collaborative diagrams.
Concurrency control in computer supported cooperative work (CSCW) has been left to social protocols, implemented as single user floor control or developed using database techniques. Since interference was observed when using only social protocols with CoDiagram, it is concluded that social protocols alone are not sufficient. CSCW systems with database style locking have had a problem with long transactions blocking user actions. A new concurrency control method, Concurrency Control for Collaborative Diagramming (CCCD) uses constraints from the diagram domain to identify user transactions. These user transactions are non-traditional transactions that supplement the traditional database transactions that maintain internal data. This improvement in user transaction identification reduces resource blocking by releasing locks without explicit user actions. It was found that CCCD would have prevented a significant amount of the interference that occurred with only social protocols.
7 Human-Computer Intereaction Diagram creation Workshop Usability and Interference for Collaborative Diagram Development Collaborative Editing Workshop at ACM Computer Supported Cooperative Work 2000 J. Campbell 2000 Campbell, Jeffrey D. (2000). Usability and Interference for Collaborative Diagram Development. "Collaborative Editing Workshop" at ACM Computer Supported Cooperative Work 2000, Philadelphia, PA. update pending
59 Curriculum Database ContinEd Web Database Development, 12 hour course University of Pittsburgh, Computer Learning Center J. Campbell 1999 Campbell, Jeffrey D. (1999). Web Database Development, 12 hour course.  Computer Learning Center, University of Pittsburgh, Pittsburgh, PA. update pending
15 Human-Computer Intereaction Diagram creation Workshop Semantic Concurrency Control for Collaborative Diagramming Consistency Maintenance and Group Undo in Real-Time Group Editors Workshop at ACM Group '99 J. Campbell 1999 Campbell, Jeffrey D. (1999). Semantic Concurrency Control for Collaborative Diagramming. "Consistency Maintenance and Group Undo in Real-Time Group Editors Workshop" at ACM Group '99, Phoenix, AZ. update pending
60 Curriculum Programming Continuing Education Visual Basic Database Programming, 18 hour course University of Pittsburgh, Computer Learning Center J. Campbell 1998 Campbell, Jeffrey D. (1998). Visual Basic Database Programming, 18 hour course.  Computer Learning Center, University of Pittsburgh, Pittsburgh, PA. update pending
5 Human-Computer Intereaction Diagram creation Doctoral Consortium Workshop Concurrency Control for Real-Time Diagramming ACM Conference on Computer Supported Cooperative Work J. Campbell 1998 Campbell, Jeffrey D. (1998). Concurrency Control for Real-Time Diagramming. CSCW '98 Proceedings of the ACM Conference on Computer Supported Cooperative Work, pp 420 - 421, Seattle, WA, ACM Press. Abstract
Diagrams represent a design, concept or object. Multiple users working together simultaneously to create a diagram can interfere with each other’s work. At a minimum this results in lost productivity. If undetected, the interference can cause inconsistencies or errors that can make the dia-gram worthless. A concurrency control mechanism is needed to maintain integrity for collaborative diagram-ming.
The method described here focuses on identifying trans¬actions for collaborative diagramming. This emphasis on transaction identification is a key distinction between this techniques and prior CSCW concurrency control ap-proaches. The improvement in transaction identification along with an implementation of split transactions reduces resource blocking, a problem generally found in applying locking techniques from database to CSCW applications.
3 Human-Computer Intereaction Diagram creation Conference Presentation of Paper A Visual Language System for Developing and Presenting Internet-based Education DOI 1998 IEEE Symposium on Visual Languages J. Campbell and D. Mahling 1998 Campbell, Jeffrey D. and Dirk E. Mahling (1998). A Visual Language System for Developing and Presenting Internet-based Education. 1998 IEEE Symposium on Visual Languages, pp 66-67, Halifax, Nova Scotia, Canada, IEEE Computer Society. Abstract
This paper describes a visual language for organizing educational materials to be published on the World Wide Web using the Collaborative Multimedia Instructional Toolkit (CoMMIT). The developer uses the visual language to create a precedence graph that defines the organizational structure of the materials. This graph specifies materials that must be viewed as prerequisites for viewing other materials. This provides significantly more flexibility than existing path-based educational authoring systems or hypertext link-based HTML editors. Additional icons representing questions and advice supplement the precedence graph to provide educational support.
2 Human-Computer Intereaction Education Conference Presentation of Paper Supporting Collaborative, Problem-Based Learning Through Information System Technology DOI IEEE Frontiers in Education Conference G. Lautenbacher, J. Campbell, B. Sorrows and D. Mahling 1997 Lautenbacher, Glenn E., Jeffrey D. Campbell, Bryan B. Sorrows and Dirk E. Mahling (1997). Supporting Collaborative, Problem-Based Learning Through Information System Technology. Frontiers in Education Conference, Volume 3, pp 1252-1256, Pittsburgh, PA. Abstract
This paper describes a suite of flexible software tools (CoMMIT-Collaborative Multi-Media Instructional Toolkit) that provides computer support for a variety of educational models including cooperative, distance, and problem-based learning. Students using CoMMIT have worked both synchronously and asynchronously in a variety of domains. The CoMMIT system provides a comprehensive, collaborative learning environment that includes integrated modules for authoring, student learning and student evaluation. This paper briefly introduces each of the three CoMMIT modules and focuses on the WWW-based student module using a problem-based learning (PBL) case to illustrate its functionality. The CoMMIT student module was recently used and evaluated against a paper-based PBL case that required students to collaborate on the design of an information system. An evaluation of the results is given
29 Other Education Undergradute Multimedia Physics Problems for the WWW using CoMMIT University of Pittsburgh J. Campbell and L.Tena 1996 Campbell, Jeffrey D. and Leonard Tena (1996). Multimedia Physics Problems for the WWW using CoMMIT. Pittsburgh, PA, University of Pittsburgh. Instruction and Learning course project
4 Human-Computer Intereaction Education Conference Presentation of Paper Visual language for web authoring Designing for the Web: Empirical Studies Conference J. Campbell and D. Mahling 1996 Campbell, Jeffrey D. and Dirk E. Mahling (1996). Visual language for web authoring. Designing for the Web: Empirical Studies Conference, Redmond, WA., October 30, http://www.microsoft.com/usability. update pending
61 Other Other Trade press Measuring Productivity Graphic Network J. Campbell 1991 Campbell, Jeffrey D. (1991). Measuring Productivity. Graphic Network. vol 1, Issue 76, 2 pages. Regional printing industry magazine.
71 Nature Humor Invited Talk 30 Solar Cycles of Turtle Research at Jug Bay, Honoring Director Chris Swarth on his Departure for California, December 14, 2012  PDF Jug Bay Wetlands Sanctuary J. Campbell 2012 Campbell, Jeffrey D. (2012). 30 Solar Cycles of Turtle Research at Jug Bay. Jug Bay Wetlands Sanctuary, December 14, 2012 Abstract
Chris Swarth researched Jug Bay turtles for many years. He was apparently unaware that the turtles were also conducting research on the biped invasive species. This talk summarized the results from the research performed by the turtles, particularly their research on the behavior of their favorite subject, "the Chris." Translated from the original turtle scratches and presented by Jeffrey Campbell.
72 Environmental Informatics Water Quality Conference Poster Analysis of Volunteer and Automated Water Quality Data: Recommendations for Improvements and Data Management, April 28 - May 2, 2014  PDF 2014 National Water Monitoring Conference J. Campbell 2014 Campbell, Jeffrey D. Analysis of Volunteer and Automated Water Quality Data: Recommendations for Improvements and Data Management poster presented at the 2014 National Water Monitoring Conference, Cincinnati, OH, April 28 – May 2, 2014 Abstract
The analysis recommends procedural improvements based on comparison of volunteer water quality data with an adjacent automated data logger. Parameters include dissolved oxygen (DO), water temperature, pH, and salinity. Air temperature was also available from a weather data logger. Based upon the time of each volunteer observation, a linear interpolation of the data logger’s prior and next observation value was computed for each parameter. Almost 150 volunteer observations were matched to the data logger since it was installed in 2003. Manual sampling had begun at this tidal fresh water site on the Patuxent River in 1985.
Volunteer and automated water temperature observations were very close. Limitations in the sensitivity of volunteer pH and salinity equipment were apparent in the comparison. There were greater differences between volunteer and data logger values in air temperature than water temperature. The differences in DO values were further analyzed over time, by season, by volunteer, by equipment and various other factors.
Procedural recommendations include entering data stored in the meter used by the volunteers for comparison to data entered from the manual data sheets, improvements to some aspects of the sampling procedures manual, and greater attention to reducing “outliers.” The plan had been to discontinue manual sampling adjacent to the data logger, but the analysis here shows that such sampling should be continued to compare individual volunteer results to the data logger to identify training needs or potential problems with the data logger.
73 Environmental Informatics Remote Sensing Tech Report Further Analysis of Maryland Offshore Wind Climatology with Application to Wind Power Generation Client project J. Campbell 2013 Campbell, Jeffrey D. (2013) Further Analysis of "Maryland Offshore Wind Climatology with Application to Wind Power Generation" Tech Report delivered to customer April 2013, publication pending, 14 pages. Description
Independent review of report assessing wind power potential using synthetic apperature radar (SAR) data to determine wind speed based on the roughness of the ocean surface. Compared prior results with nearby buoy data. Reanalyzed data to adjust for statistically non-random and non-independent satellite observations.
74 Environmental Informatics Data Management Authoring team member Harnessing the Power of Digital Data: Taking the Next Step Scientific Data Management (SDM) for Government Agencies...,  PDF CENDI Workshop to Improve SDM held June 29 - July 1, 2010 CENDI 2011 CENDI, Harnessing the Power of Digital Data: Taking the Next Step. Scientific Data Management (SDM) for Government Agencies: Report from the Workshop to Improve SDM. Workshop held June 29 - July 1, 2010, Washington D.C. March 2011. Report No. CENDI/2011-1. Co-sponsored by the Environmental Protection Agency (EPA), CENDI, and the Interagency Working Group on Digital Data (IWGDD), Washington, DC, 2011. Details
Jeffrey D. Campbell identified as Authoring Team member.
75 Nature Humor Invited Talk 15 Growing Seasons of Wild Research at Jug Bay, Honoring Education Coordinator Elaine Friebele on her retirement, November 22, 2014 Jug Bay Wetlands Sanctuary J. Campbell 2014 Campbell, Jeffrey D. (2014). 15 Growing Seasons of Wild Research at Jug Bay. Jug Bay Wetlands Sanctuary, November 22, 2012 Abstract
Elaine Friebele retired after 15 years at Jug Bay. For several summers, she led the infamous wild rice research project in the marsh. She was apparently unaware that the rice plants were also conducting research on the biped invasive species. This talk summarized the results from the research performed by the rice. Told by the rice to the turtles. Translated from turtle scratches and presented by Jeffrey Campbell.
76 Nature Humor Invited Talk Citizen science water quality compared to adjacent NERRS sensors Patuxent River Conference, July 18, 2015 J. Campbell 2015 Campbell, Jeffrey D. (2015). Citizen science water quality compared to adjacent NERRS sensors, Patuxent River Conference, July 19-19, 2015, St Leonard, MD. Abstract
128 observations of the Patuxent River water quality were compared with nearby continuous monitoring values at Jug Bay. Air and water temperatures were generally consistent but with some outliers. Differences in the methods/instruments for salinity and pH explain variability in those parameters Manual versus con mon DO values showed variability that was further investigated. Seasonality and sensor drift were not apparent. Inter-observer variability was significant. The overall results show the manual sample location is similar to the con mon. Taking a reading with a hand-held meter there could be part of the QA plan for other research at that site.

For details click on + icon in first column.
To sort click on the column heading, shift-click to sort by up to 3 columns.