The discussion ranged from skills to succeed in an organization using GIS to support business decisions (City of Seattle), ESRI, the leading GIS software producers and freelance development using GIS technologies. There is significant overlap for the required and desirable skills, but also some slight differences. I will list the skills in no significant order and provide a brief explanation. If a skill is something I personally added, it will start with an asterisk.

GIS Skills

• 
• Spatial Data and Algorithms understanding: Understand the special case of spatial data, how they work and their internals. Also, be familiar with how certain operations are carried out and when they are applicable. Many operations will run in the software, but not necessarily produce valid results. (Contributed by reader Duane Marble)
• Data entry: Be able to enter data into a database successfully with minimal errors. This includes editing said data as needs arise.
• Data conversion: The ability to convert data from either older sources (digitization) or from multiple sources to either a common format or common schema. It is extremely useful to be able to work with data coming from GPS and performing data corrections as needed. (With contribution by Jimmy Xu)
• Data maintenance: Be able to maintain data, correctly archive and ensure quality control.
• *Metadata creation and editing: Maintain logs of data processing and relevant information to include in metadata and ensure accurate creation and maintenance of said metadata.
• GIS Analysis: Be able to perform GIS Analysis as it is often used to solve common problems. An ability to extend and alter the standard analysis to meet requirements is a plus. Remember, data analysis can be performed on vector or raster data, therefore some remote sensing skills are required. (With contribution by Jimmy Xu)
• GIS Workflow: Understand the workflow to perform some procedure and be able to follow it and enhance it as needed.
• Model Building: Be able to create models of processes to allow for a workflow to be built. Also, model building in the ArcGIS sense is very helpful in this regard.
• Cartography and Graphic Design: Familiarize yourself with cartographic principles and graphic design principles. Maps are used in a variety of ways and presented in a multitude of media. You need to be able to work with that. Think of color, symbology, fonts, etc. Bad cartographic design will often make your analysis hard to decipher and interpret. (With contribution by DavidM)

Programming Skills

• Basic understanding of programming: Be able to understand what programming is and what it can do to solve certain problems. Know the strengths and limitations of programming custom solutions to problems, as well as time requirements. (More about programming)
• Programming language: Familiarize yourself with a programming or scripting language, as it is often used to build workflows or custom solutions to problems. For scripting language, both ESRI and the open source community tend to gravitate toward Python. For programming languages, C++ will give you an opportunity to work in multiple environments, while C# and the .Net languages offer you good development tools and interaction with Windows based software. (More about programming languages)
• Object Oriented programming: Learn the concepts of object oriented programming and be able to apply them in conjunction with your programming language of choice. Most GIS development is leaning toward this paradigm, and you should too. (More about object oriented programming)
• Basic GIS architecture (desktop and web): Understand the architecture of GIS and the method of communication between the different parts of GIS. Be able to distinguish when one can introduce internet-based communication in the mix and how it would work. (More about GIS architecture)
• Web Services knowledge and experience: Web services are everywhere these days, and GIS is not escaping. Learn about them, how they work, and try to implement some of your own. HTML, CSS, JavaScript, XML and related AJAX technologies are a valuable tool. (More about web services) (With contribution by Andy Anderson)

Database Skills

• Able to understand data models and structure: When given a database, you should be able to explore the data models within it and understand the structure of the database. Often times, structure will be represented in diagrams (UML), discussed below. (More about data models and structure)
• Ability to design data models: Given specific requirements for data, you should be able to design data models to fit your data.
• Database Design tools knowledge: You should familiarize yourself with database design tools, like Microsoft Visio. Most design work for data models uses it. (Check out Visio here)
• Structured Query Language (SQL) knowledge: Almost all modern Database Management Systems (DBMS) understand SQL for data queries, inputs, deletions, etc. One should be familiar with SQL and be able to perform SELECT, INSERT, MODIFY and DELETE statements. JOINS, RELATES and further SQL knowledge is greatly valued and useful on the field. (Contributed by Andy Anderson)

Project Management and Design

• Ability to translate user needs to solutions: More often than not, you will be supporting some client (or boss) that is not familiar with the details of GIS. You need to be able to translate their needs into solutions that can work in your domain. If it is not possible, you also need to be able to say so and offer alternatives. This is similar to requirements analysis used in software development.
• Good communication skills: You need to be able to communicate effectively and with confidence with your team and clients. There is no substitution for this skill.
• Good writing skills: Communicating is not restricted to verbal communication. You need to be able to clearly communicate in writing not only for communicating with your clients, but also to be able to produce metadata and reports of your work.
• Project management skills: Often overlooked in the GIS world, formal data management training is desired and required to run successful projects on time and within budget.

Other Skills

• Ability to apply expertise in multiple domains: GIS skills, while important, are not useful if they can not be applied to different domains. Your knowledge of other domains (like biology, forestry, etc) will allow you to think of creative ways to apply your GIS skills in a multi-disciplinary functions, which is greatly needed. Think outside the box (Yawer S. Ansari commented to reiterate this)
• Portability of skills on multi-platforms and online/offline world: Your skills need to be applicable to different platforms. Not only should you learn how to do GIS Analysis, but you should be able with limited help to achieve similar results using other platforms (be it moving from an ESRI training to Idrisi, or from a Windows machine to a Linux machine, or a desktop application to a server based one). An ability to traverse between online and offline worlds is a valuable asset to have.
• Detail oriented: This needs no explanation, but in the GIS world, detail oriented can get you very far. The quality of your work will show (especially when you think of metadata or workflows).
• Customer Support skills:  In most cases, GIS is used as a support tool within large organizations. As such, GIS Analysts oftentimes need to interact with clients, either internal or external. Having good customer support skills ensures you establish strong relations and opportunities.
• Don’t be afraid to explore

Now that you’ve read all the skills mentioned at the panel discussion, please provide your own or your feedback below.

UPDATE: Fixed spelling. Thank you skobola for the corrections.
UPDATE: Incorporated user comments into the body of the text by Yawer S. Ansari, Duane Marble, Andy Anderson, DavidM and Jimmy Xu.

Related posts:

1. The Essential Python Modules for GIS

Google Earth Logo

Before we begin, let us make a distinction. Google Maps and Google Earth are not one of the same. Google Maps is a web-based application that allows people to view maps on a web page. There are three main features Google Maps offers:

1. Users can search by location (i.e. for restaurants within their current map view).
2. Users can ask for directions in a multimodal transportation system (i.e. driving, public transport, etc).
3. Users can add their own data to the map, but only if they can master the Google Maps API (not a hard task, but not something the average person will do).

Google Earth on the other hand is a desktop application that provides the first two functionalities (search and direction), as well as an easy way to add your own data, one point at a time, to your map. There are of course some more capabilities present, like measurements of paths and the ability to upload and share your data with other Google Earth users. But Google Earth is not only what most people user. There are 2 more editions.

1. Google Earth Pro
   Provides some data import functionality from existing GIS data sources (shapefiles, etc) and spreadsheets.
2. Google Earth Enterprise
   The infrastructure that power Google Earth and Google maps at your fingertips, to create your own versions of them. The main advantage is data management of terrabytes of data, with 2 dimensional web maps and 3 dimensional globes.

All of Google’s offerings in the arena of geographic data are impressive to say the least, and worthy of attention. But are they truly GIS?

If one wants to define GIS, one needs to run back to definitions used throughout the years by academics and software publishers alike. Leaving the issue of people running the system and organizational contexts that build the system in a certain way, GIS is defined by the following components:

• Data: The very basic component is of course measurements. Without them, a GIS cannot function. Undoubtedly, Google has made data available in an unprecedented way.
• Representation: Measurements need to be represented in a format that the computer understands in order to operate on. Google Maps/Earth do have this. The famous KML format is one example on how Google does it.
• Operations: This is where Google Maps/Earth get a bit left behind. Granted, there is the search functionality and route finding operations, but is this the only thing a GIS can do? Out of the box, most GIS offerings have operations for slope calculation, linear referencing, buffering, and many more. Google Maps/Earth lack this functionality at their current state.
• Transformations: In this case, we do not simply refer to transformations from one data format to another, but also transformations from simple measurements out in the world, to the results of an analytical process (a report if you may). The pathway from measurement to data to information is not clear, and unfortunately not present in many cases.

If one uses then one of the commonly accepted definitions of GIS, one cannot easily place Google Maps/Earth in this realm. This is not to say that they do not partially provide functionalities that a GIS would, but rather that they are limited and not the whole picture. A pollution dispersion problem would not be solved by Google Maps/Earth. A study of the distribution of populations among some characteristic cannot be rigorously performed (visually perhaps, but can anyone report error numbers visually?).

I therefore suggest that Google Maps/Earth, as well as offerings by other vendors, like Virtual Earth, Bing Maps, Yahoo! Maps and MapQuest should not be called GIS, but rather Mapping Information Systems, or even Networked Mapping Tools. Partially a GIS, but lacking the full functionality needed.

Related posts:

1. Google Spatial Data in-house – Goodbye to TeleAtlas?

Related posts:

1. Cartographic Typefaces: Some examples

As you may notice, this map looks strikingly similar to the map for 2006. While the numbers have changed slightly, the trend remains the same for all countries, making the classification scheme used (Jenk’s Natural Breaks Algorithm) to produce a similar categorization. This will need to be address by either changing the classification scheme or normalizing the data.

No related points found for this point.

Enjoy!

No related points found for this point.

It is interesting to hear about recognition of Geography and GIS outside the discipline, how the taskforce managed to accomplish its goals and once again, the help provided by ESRI (requiring products was, according to me, a great help). Enjoy listening to it, especially if you are familiar with the NCGIA Core Curriculum Project.

Please note the following golden quote:

Academic Geographers were severely over-represented in this first edition and the reason is quite plain. Academic geographers are paid to do [...] service… My hope is that now that this effort has received some attention. [...] I think we will be able to do a good job recruiting a more deverse number of people from industry as well as academia[...]

This may actually be key to one of my upcoming posts on GIS and education.

No related points found for this point.

The question raised is whether teaching GIS in such a setting and format is ideal or not, and how can it be improved.

What is GIS?

Geographic Information Systems rely on a multitude of disciplines to build a system of “software, hardware and procedures to facilitate the management, manipulation, analysis, modelling, representation and display of georeferenced data to solve complex problems” (please see Nicholas Chrisman’s excellent lecture covering what GIS is). While this definition of a GISystem is not universally accepted, everyone agrees that it is a multi-component system. A complete and all-around curriculum on GIS needs, in essence, to tackled those multiple components on two levels: theory and application.

Teaching the Theory

In educating people about GIS, curricula throughout universities cover the theories behind GIS. From mathematical representations of data (as vectors for discrete and raster for continuous data), mathematical (geometrical) operations, to computational methods to store, retrieve and analyze data, to statistical and systems modeling techniques and even the art of representing data either as pictographs, charts or even text using visual hierarchies and the like, curricula are in need to cover expose students to a plethora of disciplines to allow them to understand what happens behind any “black-box” GIS implementation (i.e. the software application).

Teaching the Software Application

All these theories and knowledge are what the software package uses to manipulate, analyze, model, represent, display and manage spatial information. How would that then translate into teaching students how to use software. The traditional approach is to show students how operations are performed through instructional sheets and requiring students to follow directions closely, leaving little room for exploration. But most importantly, leaving little room for students to bridge the theory taught in lecture with the press of a button required by the applied portion of the curriculum.

Problems in the Current GIS* Teaching Methods

Through my own personal experience, as well as discussion with other students and educators within the realm of GIS* (GIS* symbolizing the GIScience and GISystems together), I have come to the conclusion that there is a rather large gap between the theory (or what some like to call, GIScience) and the application to perform what the theory teaches (GIS in the traditional sense). The gap seems to arise from the fact that the software system does not require any prior knowledge of its inner workings (and hence, theories) to run, while the curriculum requires students to learn both, simultaneously.

The problem is, therefore, a disconnect between theory and application, lecture and lab, what the educator feels is needed to be taught (the theory) and what the students feel they need to learn (the software).

Solving the Problems

How one can proceed to solve the problem stated above depends on whether there is an agreement that this is indeed and the case, and whether one believes that this is a problem. Universities are institutions that mainly want to push a more academic agenda (as theory is born and raised within universities) than what is sometimes needed in the workforce. An emphasis in theory is therefore understandable, even desired (if no one develops and challenges new theories, there would be no progress). But the need of students is (usually) to learn the software in greater detail, as it is unlikely that outside a university setting they will need to recite the fundamental levels of measurement (nominal, ordinal, interval, ratio according to Stevens, or adding graded memberships, cyclic ratio, derived ratio, counts and absolute according to Chrisman). How can then one proceed, especially when one considers the student body which includes students from multiple fields and with different needs?

I have pondered the question multiple times so far and I have not yet come into any conclusion. But I do believe there is a need to change the way things are taught, either by separating classes into theory and application, or changing the current curriculum to look more closely into the issue of what the students want and need to learn, as well as what the workforce the University prepares them for needs.

Notes:
A Very Spatial Podcast Episode 77 covers the issue as well.

No related points found for this point.