Quality communication in social media

no caption required

As a professional statistician, I help doctoral students design quantitative research studies, and analyze and interpret their data, for their dissertation.  It occurred to me that some aspects of a doctoral dissertation could be applied to social media communications to make information and conversations more rigorous and tractable.

If all you have is a hammer, is every problem a nail?

My answer:  No.

If all you have is a hammer, it simply means you are only going to be able to solve problems that involve driving nails.

My point is this, as a professional statistician, I help doctoral students design quantitative research studies, and analyze and interpret their data, for their dissertation.  Thus, I tend to think about research problems from a quantitative perspective.

Although many research problems cannot be studied quantitatively (e.g. we may not yet know how to quantify the constructs of interest), many of them can.

When a problem can be discussed from a quantitative perspective, why not discuss it in a format like a miniature dissertation?

I have read thousands of doctoral dissertations, and I have personally helped hundreds of doctoral students with the statistical aspects of their dissertation.  So, my perspective might be a little like the hammer and nail analogy.

I’m used to studying research problems in a dissertation format, and maybe I think that will work well in other areas, like in social networking dialogs, when in fact, I am just seeing the problem (quality of communication in social networking mediums) as a nail that my hammer will work on. But, let me throw this out there and you decide.

The typical dissertation I work on has five chapters:

1. Introduction;
2. Literature review;
3. Methodology;
4. Results, and;
5. Discussion

So, how might we apply this to a post? Let’s pick the one titled:  Project Communication and Social Networking. Toby made a comment that produced some interesting dialog and debate about project management success rates.  Toby claimed: “the majority (65% – 90%) of all projects fail – organizations don’t know how to scope and manage projects”.

Another blogger disagreed with that statistic.

How might Toby have made his case in a brief dissertation format?

Let me take a shot at it…

Introduction
Project failure could be defined a number of ways, such as failure to complete the project at all, failure to complete the project on time, or failure to complete the project within budget. Project failures rates can be attributed to factors relating to strategy and/or project management.

Literature Review
According to source 31% of projects are cancelled before they are completed and 53% of projects come in at 189% of the original budget.  Several other studies on that web site show that Information Technology project failure rates vary from 36% to 80%.

Methods
A Google search was conducted in an effort to identify published statistics on project failure rates.  The following search term was used for the search: Standish Project Failure
Google returned over 19,000 search results. Only the first two were inspected.

Results
Based upon a very limited Google search, and very limited time spent reading articles on project failure rates, there is some evidence to suggest project failure rates in the Information Technology industry may be somewhere between 36% and 61%.

Discussion
Based upon this very limited study, there is some evidence to suggest Toby’s estimates are accurate.

Further study might include:

1. Spend more time reading the articles listed on the web site in the Literature review section;
2. Read more of the search results produced by Google for the search phrase: Standish Project Failure;
3. Check the validity of the oft-cited Standish Report (.pdf link);
4. Choose other search phrases to identify more articles on project failure rates;
5. Perform similar searches using academic journal databases.

This is not a perfect example by any means, but I thought it would be worth shooting this out there to see what others think.

I guess the point I am making is, the quality of communication in social networking may improve if writers will make a mini dissertation out of their post.  Give the reader a brief background about the subject; cover the lay of the land so to speak.

The background identifies some problem that has social implications and your post sets out to say something about it.

You would like others to take you seriously so that you can get some interesting feedback, so:

1. Back up your claims with some reputable sources;
2. Discuss the limitations of your perspective and supporting documentation, and
3. Make recommendations for how to better address the problem.

Today’s guest blog was written by Steve Creech of Statistically Significant Consulting.  Steve specializes in handling all of the statistical aspects of doctoral dissertation research, from developing their research questions, hypotheses, survey design, data analysis plan, power analysis and sample size justification, and performing the statistical analysis of their data.

I am happy to welcome Steve’s thoughts and experience as a guest contributor. Please find more insight and thoughts from Steve through his professional biography, web site, or on Twitter @StatisticsLLC