Which Counties Are in Which Regions?

You'll see in the Brogan Survey that we've divided responses into regions so we can look at geographic differences in how people respond. (Our divisions largely mirror the national DMAs, or Designated Market Areas, assigned by Nielsen as television markets. That's why the "Charlotte" region goes all the way up to the Virginia border, for example--because the Charlotte TV market does.)

Here is the list (NC first, MI second) of which counties are in which regions (the number after the region name is how many of the 600 survey respondents come from that region, and that's proportional to the population):

NORTH CAROLINA
REGION ONE/ WESTERN NORTH CAROLINA (50): Buncombe, Cherokee, Clay, Graham, Haywood, Henderson, Jackson, Macon, Madison, McDowell, Mitchell, Polk, Rutherford, Swain, Transylvania, Yancey

REGION TWO/ CHARLOTTE: Alexander, Anson, Ashe, Avery, Burke, Cabarrus, Caldwell, Catawba, Cleveland, Gaston, Iredell, Lincoln, Mecklenburg, Richmond, Rowan, Stanly, Union, Watauga

REGION THREE/ PIEDMONT-TRIAD NORTH CAROLINA: Alamance, Alleghany, Caswell, Davidson, Davie, Forsyth, Guilford, Montgomery, Randolph, Rockingham, Stokes, Surry, Wilkes, Yadkin

REGION FOUR/ SOUTHERN NORTH CAROLINA (39): Bladen, Brunswick, Columbus, New Hanover, Pender, Robeson, Scotland

REGION FIVE/ TRIANGLE REGION (176): Chatham, Cumberland, Durham, Edgecombe, Franklin, Granville, Halifax, Harnett, Hoke, Johnston, Lee, Moore, Nash, Northampton, Orange, Person, Sampson, Vance, Wake, Warren, Wayne, Wilson

REGION SIX/ COASTAL NORTH CAROLINA: Beaufort, Bertie, Camden, Carteret, Chowan, Craven, Currituck, Dare, Duplin, Gates, Greene, Hertford, Hyde, Jones, Lenoir, Martin, Onslow, Pamlico, Pasquotank, Perquimans, Pitt, Tyrrell, Washington

MICHIGANNORTHERN MICHIGAN: Alcona, Alger, Alpena, Antrim, Baraga, Benzie, Charlevoix, Cheboygan, Chippewa, Clare, Crawford, Delta, Dickinson, Emmet, Gogebic, Grand Traverse, Houghton, Iosco, Iron, Kalkaska, Keweenaw, Lake, Leelenau, Luce, Mackinac, Manistee, Marquette, Mason, Mecosta, Menominee, Missaukee, Montmorency, Newaygo, Oceana, Ogemaw, Ontonagon, Osceola, Oscoda, Otsego, Presque Isle, Roscommon, Schoolcraft, Wexford

WEST/SOUTHWEST MICHIGAN: Allegan, Barry, Berrien, Branch, Calhoun, Cass, Kalamazoo, Kent, Muskegon, Ottawa, St Joseph, Van Buren

CENTRAL MICHIGAN: Clinton, Eaton, Gratiot, Hillsdale, Ingham, Ionia, Isabella, Jackson, Lenawee, Montcalm

EAST CENTRAL MICHIGAN: Arenac, Bay, Genesee, Gladwin, Huron, Lapeer, Midland, Saginaw, Sanilac, Shiawassee, Tuscola

OAKLAND COUNTY: Just Oakland County

MACOMB COUNTY: Just Macomb County

WAYNE COUNTY: People who live in Wayne County, but not in the City of Detroit

CITY OF DETROIT: People who live in the City of Detroit

REMAINDER OF DETROIT MSA: Livingston, Monroe, St Clair, Washtenaw

Margin of Error of the Brogan Survey

The Brogan Survey is a 600-person telephone survey. As a result, the simple answer to the margin of error question is +/-4% at a 95% confidence level. If that's the answer you need, then you're all set.

Many people don't really understand margin of error, which is surprising, since Statistics classes where just ooohhhh so fascinating. The +/-4% that you see in the paper all the time is the worst case scenario for the margin of error. That's the margin when 50% of people answer a question yes and 50% answer no. So the actual split is somewhere between 50/50 and 46/54 either way.

As the population gets more united in any answer (for or against something), the margin of error decreases!

• So if 60% agree in a Brogan survey question, the margin is 3.92%
• If 70% agree, the margin of error is 3.67%
• If 80% agree, the margin of error is 3.2%
• If 90% agree, the margin of error is 2.4%
• If 99% agree, the margin of error is 0.8%

Bored yet? No? Ok, glutton, here comes your punishment...

Some people ask how 600 people can accurately represent the 8 million or so people in each state we survey in... Well, they can't entirely accurately, which is why you have margin of error in a survey. But if you jumped up to interviewing 1,100 people, the margin of error would only reduce to 3%--and the survey costs way more to conduct. And that's, well, bad, mostly for us...

But the fact is that even if you interviewed 1 million people, the numbers won't change too much either way. That is, unless you did a survey that was really poorly designed, so that all callers were very much alike. But if you do a true, random-digit dial phone survey, where all areas of the state are equally represented, and ethnic breakdown matches the state, and income breakdowns are accurate (like we do), you get a pretty good idea of what the entire state is thinking.

If you're not bored yet, you're a stronger person than I am...

Why Does Brogan Conduct the Brogan Survey?

Our philosophy as an agency can be summed up by a quote from Professor Robert Lauterborn, the chair of the advertising department at UNC-Chapel Hill, and the author of Integrated Marketing Communications, who said:

"The only sustainable source of competitive advantage is superior knowledge of the customer."

So, for our clients to win, we have to be better at understanding their customers than anyone else. That's why we invest so heavily in secondary research tools. To supplement that, we decided to conduct our own original research in our primary business markets: Michigan and North Carolina. The Brogan Survey was born out of that idea.

We have no vested interest in the answers to these questions. We're not advocating for a particular position, product or candidate. We're simply working to understand the mood of the consumer in markets that are key to us and to many of our clients.

Please let us know what you think of the survey using the comments field.

Introducing the Brogan Survey

Coming mid-May 2007, the Brogan Survey will ask consumers in North Carolina and consumers in Michigan a series of questions to gauge their attitudes on the economy, health care, business and marketing issues and current events.

The Brogan Survey is a statewide, 600-person, random digit dial telephone survey conducted by human interviewers, not robo-calls. The survey is statistically valid, with the sample stratified to ensure a representative sample of all areas of the state proportional with actual population distribution.

The Brogan Survey debuts in May of 2007 with the first North Carolina poll. The first Michigan poll will take place in June of 2007. The Brogan Survey will be conducted six times a year (three in Michigan, three in North Carolina).

We'll use this blog to share the results with the blogosphere. We hope you enjoy. Please leave a comment to let us know your thoughts.