Research, Analysis and Writing

In doing your research and writing your paper I suggest going through a linear, step by step process like the following:

  1. Find a topic or area of interest (I have several suggestions in the paper overview on blackboard.)
  2. Start researching, look through Wikipedia or another encyclopedia for the basics and then look deeper for data and other peoples’ analysis. (My links page will be helpful here.)
  3. Based on the data and information you’ve found, narrow your research question or topic if necessary and decide exactly what you’d like to write about.
  4. Think of the structure of your paper (suggestions in the paper overview on blackboard).
  5. Find data to play around with. (Your papers have a charts and data requirement that can be easily satisfied by using excel.  Knowing excel extends beyond our class and is often a skill requirement in jobs. This 30 minute youtube video provides a much better overview of using excel for basic charting, graphic and analysis than I could hope to provide in a handout, ergo I suggest you watch it a few times. When you have a question, google or youtube it, chances are someone has asked that question before.)
  6. Gather your references together and think about how to logically flow your paper.
  7. Start writing. (Often writing the body first is best, then after you see what your argument evolves into you can go back and write the proper introduction and conclusion.)
  8. Insert graphs and tables as necessary for the illustration of key points. (I find making the chart or graphic, titling and labeling it within excel and then doing “paste special/image” into the word processor works well to avoid formatting problems.)
  9. Cite as you write, it is much less stressful than going back and attempting to cite later.
  10. Construct a suitable bibliography of sources.
  11. Proofread and edit.
  12. Proofread and edit (again).
  13. Turn it in.

Following a linear procedure is useful. Sometimes it’s hard to sit down and “write a paper” but by breaking it up into smaller pieces you don’t have to “sit down and write my ****** paper this weekend instead you can just sit down and do steps 1-3. In a few days do 4-6, then 7-9, then 10-13 or whatever works. The quality you produce will be better, the process will be far less stressful and you’ll see yourself actually making progress. It works.

On Graphing, Tabling and Charting Data

The youtube video above will tell you how to make Excel tables, charts and graphics. But the question may still remain: but what do I make tables, charts and graphics for?

The answer is really is anything you see about your topic that (1) involves numbers and (2) you feel is important.

Including tables and charts is great for you as the writer in two basic ways: First it can convey complex relationships to the reader far easier than trying to describe the data with words alone. Secondly explaining the charts you’ve included tends to write your paper for you.Each table or chart you include will probably take an entire page of text to adequately explain; including them makes your writing more effective, more convincing and easier.

An Example (Click pictures for a closer look.) 

Created in excel in ~2 minutes.

On the right is a table of some numbers I tossed into excel, which just show the values of some random variable in three countries over 21 years from 1980-2000. (Technically this is called time series data.)

Now just making this data up, I really don’t understand what’s happening in this table. Similarly it would be even more difficult for me to explain to you what’s happening in this table as again, I have no idea.

I could try to compare the numbers and say that roughly over time this variable appears to be growing in the three countries, but I would have a hard time saying how different each growth rate appears to be from the others.

Hypothetical Graph

What happens if I instead if I chart this?

Well, it looks much better. But more importantly now besides just “what is this” I now have a ton of questions to ask and answer about this data we wouldn’t have clearly seen within just a table. For instance:

Clearly red > green or blue. Why?

Clearly green suffered some setback at about 1986. What was it? Also, red and blue were unaffected, why?

Clearly all three suffered some setback in 1998. What was it, and why?

Clearly all three are recovering from the 1998 setback, but red suffered far more than blue. Why?

Obviously you now have at least a page of text necessary to explain these questions. Data can write your paper for you. 

Hypothetical Bar chartHypothetical Bar chart RateLets go further and have a little fun with this by doing some extremely basic forecasting.

So to do this all I’m doing is taking the data I created from 1980-2000 and taking the average for each country. This gives me country averages in the chart on the right.

To find a rate of growth per year I take the averages generated above and divide them by the number of years (1980-2000).

This gives me the growth rates chart on the right, which you’ll notice is almost a mirror image of the preceding chart since all I did was divide each country’s average by the 21 years.

Now to do a very simple forecast of years 2000-2020; I first make the entirely unrealistic assumption that each of these countries will maintain the same average growth rate from 2000-2020 that it did from 1980-2000.Hypothetical Table Forecast

So I take the each country’s value I had before for the year 2000 and for 2001 assume it will be the value from 2000 plus whatever the growth rate was.

Thus country B in 2000 is 96.00 but at its growth rate of 2.2% in 2001 I forecast it will be 98.11, which is simply 96+(96*0.022). The same formula is applied to the others countries and years as well, which excel can automate for you.

But again a table, while nice to include to indicate your data exists, doesn’t tell us much. Visualizing is better.

Hypothetical Chart ForecastWhat does this forecast look like graphically?

Now I’m only trying to illustrate ideas here, but you can see why forecasts run into trouble. Simply compare this to the data charted from 1980-2000 and its clear that our 2000-2020 projections look nothing like a boom or bust cycle.

Realistically you would never do a 20 year forecast like this in the real world for the obvious reason that it’s not realistic for long periods of time. But if you only forecasted 1 or 2 years (say 2001 & 2002), this simple forecast done with something like GDP or per capita GDP would generally be accurate unless there is some huge economic event in 2001 or 2002 which perturbs the data.

So basically we’ve identified a trend in a variable and are now using that trend to estimate future values of that variable. It may work, it probably won’t, but nonetheless people will pay you to do it (see: consulting, finance, marketing research, etc) because even though it rarely works, when it does work people using the forecast make money. (This also is a big piece of what econometrics is.) 

Hopefully this gives you some idea as to why charts and graphics are useful. The right data will tell you what questions you should ask and answer. Explaining and analyzing is usually easy, often the hardest part is figuring out what to analyze and explain; luckily having data just gives that to you. 

As always, don’t hesitate to contact me for help.


Grammar links to The Oatmeal:

i.e. and e.g. 



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s