1 week day and 2 weekend days Saturday and Sunday Why spread it out over

# 1 week day and 2 weekend days saturday and sunday why

This preview shows page 11 - 14 out of 16 pages.

9pm over 3 days. 1 week day, and 2 weekend days (Saturday and Sunday). Why spread it out over different days and times of the day?• Survey method – intercepting customers in the casino lobby• Sampling technique - stratified samplingThe three segments are local gamblers and domestic tourists and overseas tourists.The segments will be identified through a survey questionThe second stage probability sampling is done through systematic sampling – a researcher would stand outside the casino and ask the first 15 customers in every hour to fill out a survey• Sample size – 500 to be split equally between the three segments (small scale problem-identification research)Week 11 Causal Research (experimentation)
What is causal research (in its most elemental form)?Definition – the process of manipulating the independent variable and measuring its effect on the dependent variable, while controlling for the effect of everything else并测其对影响，同时制其他因素的影响Bottom line – you need to offer X at different values and obtain the corresponding values of YTwo ways to do thisOption 1- Create sub groups using your sample and apply different treatments (X) to different groups– Example: show different graphic designs (X) to different Facebook users and measure the click-through rate from these groups (Y)– Test for causality: are users who receive the story with the tiny flag on average more likely to click on the story than those who received the one without the flag? (typically done using a t test)• Option 2- Apply different treatments (X) over time– Example: send the standard design (no flag) (X) to Facebook users in Q1 and the one with the flag (X) in Q2– Test for causality: is the average click-through rate higher in Q2 than Q1? (typically done using a t test)

#### You've reached the end of your free preview.

Want to read all 16 pages?

• Two '16
• representative, secondary data