How can replication help researchers avoid persistent Type I and ...

[Chloeellawright]

How can replication help researchers avoid persistent Type I and Type II errors.
 
  What will be an ideal response?

Question 2

Describe the replicated interrupted time-series design that assessed the effectiveness of New York's bottle return law. Why can we use it to form a tentative causal link between the law and the reduction in some litter?
 
  What will be an ideal response?

[af]

Answer to Question 1

Replication helps avoid persistent Type I errors because a single study may result in a finding that is not reliable and that will disappear in further research. With replication, we are not likely to be victimized by random error that led to a spurious result. Replication helps avoid persistent Type II errors because a single study may erroneously miss a relationship between variables in a single study; with replication, the relationship is likely to emerge.

Answer to Question 2

The replicated interrupted time-series design monitored the amount of returnable and nonreturnable litter before and after New York mandated that soft drink and beer bottles have a deposit that is refunded on return. The investigators looked at the amount of litter in New York City and in nearby New Jersey, which didn't have such a law.
The results indicated that people removed the litter for which they could get a monetary refund in New York; the same kind of litter was left on the ground in New Jersey. In both states, the amount of litter that was nonrefundable remained unchanged.
This procedure allows for some assessment of causation because you would expect both New York and New Jersey to have comparable litter if the bottle return law had been ineffective. In New York, with the law, there was less litter. The researchers also looked at the trend over time. They discovered that the pattern was consistent with the bottle return law as actually leading to less litter.

[Chloeellawright]

TYSM

[bigcheese9]

Excellent