Basic Quantitative Research Methods for Urban Planners

by Reid Ewing

First, if you collect the data yourself, it is called primary data. It can come from a survey or an interview you create and conduct, from direct observation of human behavior, or from measurement of the environment, such as sidewalk widths or the number of street trees. If you use data that someone else has collected, you are using secondary data.

Disaggregate data simply means looking at individuals’ characteristics, such as race, income, or education level, plus individuals’ behavior, such as fruit and vegetable consumption or minutes of moderate exercise. Aggregate data is a summary of disaggregate data, meaning that information of individuals is compiled and aggregated.

Aggregate data raise issues of aggregation bias and the ecological fallacy, but may be appropriate if you are interested in differences from place to place (Ewing et al., 2017).

The unit of analysis is the entity that is being anal...

Disaggregate data enables you to conduct an analysis at the individual level, and an individual-level analysis requires ...

study may have a unit of observation at the individual level but may have the unit of analysis at the neighborhood level, drawing conclusions about neighborhood characte...

Cross-sectional data is collected at a single...

Longitudinal data is collected over time—you have at least two wa...

series data, implying many waves of measurem...

Repeated cross-sectional data is collected on the same set of variables for multiple time periods but with d...

Panel data represents the gold standard for planning research, since the repeated observation of the same cases (individuals or groups) inherently controls for individual and group differences that cannot be measured but can be assumed invariant over time.

There are four scales of measurement. First, nominal scale categorizes cases with no numerical equivalent (e.g., male/female, introvert/extrovert, city/suburb). Second, ordinal measurement can rank-order the cases, but the distances between attributes have no meaning. For example, five-star hotels are supposed to be better than lower-ranked hotels in terms of accommodations and services. However, we do not know if the distance from 5-star to 4-star is the same as 4-star to 3-star. A better example for planners is the categorization of people as low income, medium income, or high income. Thirdly, in interval measurement, we can assume equal intervals between values. Typical examples include temperature and IQ. When we measure temperature (in Fahrenheit), the distance from 30 to 40 is the same as the distance from 70 to 80. Note, however, that in interval measurement, ratios don’t make any sense and there is no true zero (0 degree Celsius does not mean no temperature). Sprawl metrics developed by Ewing and Hamidi (2014) are measured in terms of standard deviations above and below the mean, but there is no true zero when it comes to compactness or sprawl. Lastly, ratio measurement always has a meaningful absolute zero, i.e., zero indicates the absence of variable measured. Weight, age, and frequency of behaviors are examples. It is called ratio scale because you can construct a meaningful fraction or ratio with its values: someone can weigh twice as much as another, or can wa...

Variables with nominal or ordinal measurement are called categorical, whereas those with interval or ratio measurement are called continuous.

theory is an explanation of why things occur. In the preceding research question, a theory could be “if highway capacity increases, people will drive more” (“induced demand theory”). Derived from theory, a statement that is testable is called a hypothesis. A hypothesis that follows from the preceding theory is: as highway lane miles increase, vehicle miles traveled (VMT)...

In general, a theory consists of constructs or abstract concepts and how they are related to one another. Highway capacity, travel, growth, and sprawl are examples. On the other hand, a hypothesis is phrased in terms of variables, quantities that vary, are measurable, and partially capture constructs. VM...

A theory or hypothesis, or multiple theories or hypotheses, can be verbally or visually represent...

Two types of variables are defined in most studies. They are independent (or predictor, explanatory, or X) variables and dependent (or response, outcome, or Y) variables. Dependent variables represent outcomes or consequences that you try to understand. Independent variables are those hypothesized to influence or predict the values of the dependent variables.

Correlation refers to a statistical index that describes how strongly variables are related to one another. It simply means that two things happen together: For example, more highway construction is accompanied by more traffic, but we don’t know which one is the cause and which one is the effect, or whether there is a third variable that is causing both, such as population growth. Please remember that correlation does not imply causation

A best-fit model, or fitted model, has just the right number of predictors needed to explain the data well. Fitting a model is a trade-off between parsimony and explanatory power (i.e., goodness of fit), because high parsimony models (i.e., models with few parameters) tend to produce a worse fit to the data than low parsimony models.

Released in 1968, SPSS is a classic, leading software package for quantitative research in the social sciences. Thanks to its user-friendly interface and easy-to-use drop-down menus, it is a useful tool for non-statisticians as well. One of the advantages in terms of learning is its similarities to Excel, software that many students are already familiar with. Other benefits include official support and extensive documentation. Thus, SPSS is by far the most common software used in academic research generally and in planning programs.

Non-technical writing may employ a complex chronology or obscure logic, but technical writing never will. As Daryl Bem explains in Writing the Empirical Journal Article: “It is not a novel with subplots, flashbacks, and literary allusions, but a short story with a single linear narrative line” (2002, p. 4).

Non-technical writing may employ colorful imagery and emotionally charged language. On the other hand, even if the author is a strong advocate of certain ideals, technical writing “aims to be clean, clear, and unemotional” (Katz, 2009, p. 3).

A third difference concerns evidentiary standards. Peer-reviewed journal articles, in particular, must show extensive evidence in support of their arguments (Katz, 2009). Non-technical writing, on the other hand, does not have set standards for what constitutes appropriate evidence. This does not mean, h...

Finally, good technical writing involves layers of expertise. Bem states that a person who is “intelligent... (but) with no expertise... should be able to comprehend the bro...

Budinski suggests that, after identifying your underlying reasons for writing, your initial plan of action should include three basic elements: identify your audience, determine the scope of your document, and specify your particular objectives (2001, p. 56).

Budinski explains, specific project objectives (outcomes) are different from the overall purpose (intention) for writing (2001, p. 63). If you are writing a plan, as just described, the purpose is to take stock of the community and to craft a future vision. Goals and objectives might be to increase transit-oriented development, to improve access to open space, to reduce water pollution, etc.

In these circumstances, it is important to avoid ambiguous terms, loaded language, slang, and colloquialisms (Rubens, 2002).

A good technical paper should explain terminology in words that are understandable to the reader and should provide definitions if necessary. Acronyms are particularly common in planning, and need to be defined when first used.

Together, tense and perspective also influence whether your writing is in active voice or passive voice. As the name implies, writing in the active voice connects a clear subject with a clear action. The preceding Flyvbjerg quote—“I decided to study ho...

Suppose that Flyvbjerg had wished to obscure his role in the research process. In this case, he might have written, “the influence of rationality and power on planning was the subject studied.” In this passive sentence, the active transitive verb phrase (decided to study) has become a past participle combined with the verb “to be” (was st...

become the subject of the passive sentence (the influence of rationality and power on planning). And, most importantly, the subject of the active sentence—Bent Flyvbjerg, the person who will actually conduct the study—has disappeared entirely.

A good sentence should state one or two ideas clearly and simply. Here are a few tips for writing strong sentences: • Use short, simple sentences (Rubens, 2002). • Ordinarily, place a verb immediately after a grammatical subject. When the subject is separated from the verb, the reader is challenged to understand what the sentence is all about (Gopen & Swan, 1990). • Use repetition and parallel construction, within and between sentences, to improve flow, cohesion, precision, and understanding (Kolln & Gray, 2009). • Scrutinize every word. Can you substitute or remove words to make a sentence convey the same meaning more swiftly? (Ross-Larson, 1999). Especially in technical writing, sentences stripped to their “cleanest components” will generally serve you best (Zinsser, 2001, p. 7).

Try to make each paragraph have one main point (the fruit), with each sentence (the orange slices) building the overall argument.

A narrative paragraph: A narrative paragraph tells a story that the reader can relate to. Generally, it is written in chronological sequence and can contain dialogue. A planner may use a narrative paragraph to describe the general vision for a town or city, using a story to establish the goals and values of the residents.

A comparison paragraph: This type of paragraph focuses on developing the similarities between subjects. Writing a comparison paragraph is straightforward. Decide which aspects of the subjects to compare, and write down how each of these aspects is similar.

A contrast paragraph: As opposed to a comparison paragraph, a contrast paragraph details the difference between two topics. These paragraphs are particularly effective when writing to persuade or to describe. The contrast can help the reader understand that the differences between two subjects are significant and important.