Question

What kind of research design (survey etc.) would you use in the following cases?

First, for each case, indicate the independent variable (IV) and dependent variable (DV – there may be more than one).

Second, for each case, decide on a research design. If you decide on a qualitative research design, justify your choice of this design vis-à-vis a quantitative one, and also your choice of the particular qualitative method (e.g. focus group) chosen. If you decide on using a quantitative design such as a survey or an experiment, justify your choice vis-à- vis a qualitative design. Then, explain how you would manipulate or measure the IV. In the case of a survey, create at least one question to measure the IV, and at least 2 questions that can be summed up to measure the DV.

Explain in each case how you would analyze the data and test for statistical significance.

(a) the objective is to see if perceptions of vodka quality are affected by the label claiming it is made from iceberg water

(b) the objective is to establish if consumers who use electric bikes on bike paths have higher incomes and are older than the consumers who ride regular bicycles and the consumers who ride skateboards

(c) the objective is to understand why some drivers behave aggressively (honking, making rude gestures, tail-gating) when traffic is heavy

Answer 1

(a) the objective is to see if perceptions of vodka quality are affected by the label claiming it is made from iceberg water

Ans

Sample can be collected from people drinking vodka. Data from the sample is to be collected twice after some changes. Firstly, (if possible) remove the label and take data and secondly data is to be collected by having the label on the bottle. Demographic data can be collected such as name, age, gender, contact info, profession, since we would require these people again. The number of vodka drinks should be counted. Sample of 50-100 is good enough for analysis.

For example, dataset should look like this

Sr No	Number of vodka drinks (Before label)	Number of vodka drinks (After label)
1	1	2
2	0	1

Paired t-test can be used for statistical analysis.

(b) the objective is to establish if consumers who use electric bikes on bike paths have higher incomes and are older than the consumers who ride regular bicycles and the consumers who ride skateboards

Ans

Two different samples needs to be collected for this objective. Randomly people using bikes and bicycles are to be considered irrespective of their age and income. Various random sampling techniques can be utilized like Random, Systematic, Convenience, Cluster sampling. So there would be two groups one using bikes and another using bicycles. Other than usual demographics, age and income of the person is to be collected.

For example, dataset should look like this

	Bikers	Bicycles
Sr No	Age	Income	Age	Income

Two sample t-test can be used for attaining the objective depending equal or unequal variance.

Also this test needs to be used once for checking age and another time for income.

This test compares the mean from both the sample.

(c) the objective is to understand why some drivers behave aggressively (honking, making rude gestures, tail-gating) when traffic is heavy

Ans

Since the objective is “why the people behave regressively”, so a qualitative approach is suitable.

Also we will get the reason behind such behavior. Start the survey by selecting a sample which self drives and collect general demographic questions including age, gender, income, and daily travel time.

The questions can be included are:

1. How often do you honk, make rude gestures or tail-gating in a day? (Numerical ans for each)

2. Do you honk, make rude gestures or tail-gating when signal is red? (Y/N)

3. Do you honk, make rude gestures or tail-gating when signal is Green? (Y/N)

4. Do you honk, make rude gestures or tail-gating when in hurry? (Y/N)

5. Do you honk, make rude gestures or tail-gating when not in hurry? (Y/N)

6. Does your mood have an effect on honking, making rude gestures or tail-gating? (Y/N)

7. Does your travel partner have an effect on honking, making rude gestures or tail-gating? (Y/N)

8. Does your travel duration have an effect on honking, making rude gestures or tail-gating? (Y/N)

Such questions will help to attain objective. Also age, gender and income effect can be checked across these questions. Chi-Square test of independence can be used for this purpose by making appropriate groups of age, gender and income