GEOG 110 Population and Place: Investigating Place through Quantitative Data Pt2

Boundary Creation and Data Limitations

Boundaries like mesh blocks are randomly drawn to meet population thresholds, often following roads or rivers. As users of secondary data, we have limited control over boundary creation.

Deprivation measurement changes over time (e.g., landlines vs. internet access).

Overarching Aims

Quantitative geographers aim to describe and explain spatial patterns in areas like New Zealand or Christchurch.

Univariate vs. Multivariate Data

  • Univariate data: One variable (e.g., millennial boomer maps).
  • Multivariate data: Multiple variables (e.g., deprivation index with nine components).

Deprivation Mapping

Chloropath mapping uses color to represent data; dark reds indicate more deprived areas, while dark blues indicate less deprived areas. This reveals the uneven spatial distribution of factors, and informs policy interventions.

Modifiable Areal Unit Problem (MAUP)

A consequence of data collection methods, where patterns change based on the scale of geography used:

  • Scale effect: Moving from small to larger areas hides heterogeneity.
  • Zoning effect: Differently delineating boundaries, even with the same underlying data, yields distinct patterns.

Ecological Fallacy

Occurs when aggregate statistics are used to make inferences about individuals within a place. Conclusions about a place may not apply to all individuals in that place.

Mobile Phone Data

Mobile phone data shows population movement patterns in specific locations over time. It provides insights into areas' personalities based on movement patterns, such as commuting or seasonal tourism.

Applications of Quantitative Geographic Data

  • Population: Ethnicity distribution.
  • Health: Life expectancy patterns.

Life Expectancy

Life expectancy varies across New Zealand. Displaying data at aggregated levels hides underlying variation. Mapping requires balancing detail (individual data) with simplification (broad comparisons).

Exploring Relationships

Charts can illustrate relationships between variables:

  • More income inequality correlates with more drug use.
  • More income inequality correlates with less trust.
  • Lower median income correlates with higher smoking rates.

Correlation Coefficient

Denoted by rr, it measures the strength and nature of a relationship between two variables.

  • Ranges from −1-1 to +1+1.
  • −1-1: Strong negative relationship.
  • +1+1: Strong positive relationship.
  • Close to zero: Little to no relationship.

Interpreting rr Values

If r=+1r = +1, there is a perfect positive association.

If r=−1r = -1, there is a perfect negative association.

If r=0r = 0, there is no association.

Key Points About Correlation

  • Does not imply causation.
  • Some correlations are spurious.

Exploratory Spatial Data Analysis

Seeks to understand causal relationships, like why certain areas have more poverty. It may employ statistical tools to see potential relationship to help figure out what we might do about those relationships.

Government Policy & Smoking

Government policy focuses on chronic conditions (cancers, cardiovascular diseases) and modifiable factors (alcohol, tobacco, nutrition).

Interpreting Smoking Rate Maps

Different colors represent varying smoking rates; yellows indicate higher rates around 12% or more. Focus is on achieving a smoke-free target of 5% or less by 2025.

Association Between Income and Smoking

A correlation coefficient of −0.74-0.74 indicates a potential relationship where higher median income is associated with lower smoking rates, and vice versa.

Summary

  • Observe spatial patterns using mapping.
  • Use correlation to explore relationships.
  • Correlation is a starting point, not proof of causation.
  • Causation requires longitudinal data.
  • Mapping as geographical detective work.