STUDY UNIT 2b QUANTITATIVE RESEARCH METHODS Analysis of Quantitative Data

Analysing Quantitative Data: Inferential Statistical Concepts

  • Statistical significance - The scientific method requires us to go beyond describing our data—we need to test our hypotheses and make inferences about our data.
  • To do that, we need to use inferential statistics (P value/ANOVA/etc.)

Organising Quantitative Data for Analysis Purposes

Describing Quantitative Data: Descriptive Statistical Concepts

  • The first step in data analysis is to describe your data using descriptive statistics.
  • some common descriptive statistics includes:

Quantitative Report and Proposal Writing

  • Title,
  • Abstract,
    • Introductory sentence.
    • Summary of methods
    • Prediction and description of results
    • A conclusion that states the support of hypothesis and for future work
  • Introduction,
    • Research topic
    • Literature review
  • Method
    • Participants
    • Research design
    • Material
    • Procedrue
    • Data analysis
  • Results
  • Discussion

1. Entering the data

  • in table form on a spreadsheet for example.

3. Cleaning the Data

  • Ensure that there are no missing data
    (i.e. participants did not submit a complete form)

2. Coding non-numerical data

  • I.e. assigning a digit to the data so that we can graphically present it

4. How to deal with incomplete data

  • One possible way is to exclude that participant from our data analysis (list-wise deletion, i.e., delete the entire row of that participant’s information)
  • Other ways include replacing the missing data with estimates

Frequency counts (or distribution)

  • This gives you the counts and the distribution for category of a variable.
    • For example, you might have 90 males and 10 females in your study. In terms of the frequency distribution of gender in your study, you can see that it is heavily skewed towards males.

Measures of central tendency

  • These measures help you to describe the centre of the frequency distribution. Includes:
    • Mean (Average)
    • Median (Mid point of the distribution)
    • Mode (The most common or frequently occurring number)

Measures of variation

  • These measures help you to describe the variability of the frequency distribution around its centre. Includes:
    • Range = Maximum Value – Minimum Value
    • Standard Deviation (SD): Average distance between all the scores in the
      distribution and its mean

Statistical Inference

  • Goal is to statistically affirm:
    • Can we infer results taken from a sample to the population? •
    • How likely is it that these sample effects reflect true population effects versus random error?.

Null Hypothesis Significance Testing (NHST):

  • This method is used to determine the probability that sample data are good estimates of population data.

NHST Process

  • The NHST process is described as follow:

Assume Null Hypothesis is true

Conduct Appropriate Statistical Test

  • Selecting an appropriate statistical test depends on the following
    • Type of research question
    • Number of variables and level of measurement
    • Research design: e.g., between-subjects vs. within-subjects; experimental vs. non-experimental

Identify p-value

Interpret p-value

  • If the p-value is low (less than .05):
    • Reject Null Hypothesis
    • This suggests that the likelihood of obtaining such a relationship is less than 5% if there truly is no such relationship within the population
  • If the p-value is high (higherthan .05):
    • Accept the Null Hypothesis
    • This suggests that the likelihood of obtaining such a relationship is greater than 5%, if there truly is no such relationship within the population

Decide whether or not to reject H0

  • What happens if you get a non-significant finding (i.e., high p-value)?
  • Replicate study (again and again)
  • Increase the power of study by increasing the following:
    • Level of statistical significance
    • Sample size
    • Sensitivity of measurement