Samples from specific situations are not likely to create representative random samples.
Call one number from each page of an online phone listing for your town between noon and 2 p.m.
Not likely; this samples the population of people with phone numbers listed online that are home midday.
Also, we only get responses from those who agree to participate.
Survey each person leaving a local grocery store.
Survey each person leaving a local movie theater.
Walk around downtown and survey every fourth person you see.
Could you make a representative sample by surveying a few people from each of the situations described in parts (a) through (d)? Explain, and consider what problems might still remain.