Investigative Questions and Data Collection — Worked Solutions

Collect weekly laundry spending in dollars per resident. The parameter is the mean weekly laundry spending of all current residents; the analysis goal is a confidence interval, a plausible range for that mean. The population for the conclusion is all current college residents, not future residents or all college students.

No. Only 30 of the 180 members supplied information. A census of the membership records information from all 180 members; complete response within a sample does not make the sample the population.

Units are seedlings. Factors are soil type (ordinary, nutrient-rich) and daily light duration (6,10 hours). The four treatments are ordinary/6, ordinary/10, nutrient-rich/6, nutrient-rich/10. Response is seedling height after four weeks. Conditions are imposed by the researcher, so the study is an experiment.

P is prospective: drivers are selected now and data are collected now and into the future. R is retrospective: selected drivers’ past data are gathered. Both are observational because driving habits are not imposed, so neither is an experiment.

The survey is observational, collecting human data with standard questions without assigning bicycle ownership. Age is a plausible confounder because it is related to both ownership and fitness and supplies an alternative explanation of their relationship. Recorded shirt colour is not a confounder here. The observational association therefore does not establish a causal ownership effect.

A uses random selection from the city population, supporting population descriptions of sleep, but it did not impose routines and does not support a causal routine claim. B uses random assignment, supporting a comparison of routine effects under the experiment’s conditions. Its volunteer recruitment does not justify generalization to the entire city; scope is participants or similar individuals.

Collect usual trip time in minutes, mode category, and late/not-late status for each selected student. Random selection supports generalizations about current Urbanopolis students if collection is sound. Because transport modes are not assigned, mode-versus-time or mode-versus-lateness relationships are observational and cannot establish a causal mode effect.

Questions can be: Is the population mean trip time different from 20 minutes? Is that mean greater than 20? Is the population proportion arriving late below 0.10? Are mode and lateness associated, meaning not independent, among current students? What range plausibly estimates the population mean trip time? The first two concern the population mean with not-equal and greater alternatives; the third concerns the late proportion with a less alternative; the fourth concerns population categorical association/departure from independence; the fifth estimates the population mean with a confidence interval.

S is a sample survey and observational study: standard questions collect human data without imposing transport treatments. It is not an experiment and not a census of 2,400 students. Recording information from all 2,400 current students would make it a census.

Each volunteer student is an experimental unit, also a subject or participant. Factors are practice method with game/flashcards levels and duration with 10/20-minute levels. Treatments are game/10, game/20, flashcards/10, flashcards/20. Words recalled one week later is the response, measured after the assigned practice. The study is experimental because the researcher imposes these combinations.

Prospective: select students now, record current trips, and track next term’s lateness without assigning travel modes. Retrospective: select current students now and gather their previous term’s lateness and travel logs. Both remain observational because travel conditions are not imposed.

Distance is a plausible confounding variable: farther students tend to use different transport modes and also have longer travel times. These links to both the explanatory variable (mode) and response (time) provide an alternative explanation of a mode-time association, reducing support for a causal mode claim.

S samples by chance from the current-student roster and can support population inference to that group. V consists of volunteers from a cycling club and has neither random selection from all students nor support for all-student generalizations; interpretations are limited to participants or similar individuals. The teaching experiment also recruits volunteers, limiting population scope, but random assignment supports treatment-effect comparisons under the study conditions.

The parameter is the current population mean collection time μ, in minutes, for the service’s collections under the changed route. The alternative is μ≠7 minutes: the purpose asks about any change, not only an increase or decrease. In words, the current average collection time differs from 7 minutes.

Let p be the population proportion of eligible city residents who use the bike-pass service this year. The alternative is p>0.40: adoption has increased beyond the previous 40%. A sample percentage is an estimate of p, not the parameter itself.

Let p be the population proportion of packages lost under the current new-label process. The alternative is p<0.06: the current package-loss proportion is below 6%. The relevant quantity is a proportion of packages, not their mean weight or a sample count.

The alternative is that membership tier and preferred format are associated in the population of current library members: they are not independent. Equivalently, the population proportions preferring print/audio/digital are not all the same across the two tiers; at least one conditional format proportion differs. This concerns a population categorical relationship, not an alternative about a mean or necessarily a causal tier effect.

The experimental unit is one test box. The sole factor is packing material; its three levels are shredded paper, foam pieces and air pouches. Each level is one treatment because this experiment has only that one factor, so there are three treatments, not nine combinations. The response is the number of damaged items in a box after the imposed drop. The fixed drop height is a controlled feature, not a second varied factor.

This is a deliberately chosen sample of the most frequent weekday visitors, not a random sample or a volunteer sample. Their responses describe those 20 selected members, and may be useful for understanding similar frequent weekday visitors, but the design does not establish representation of all 900 members. Frequent weekday visitors may have different weekend preferences from infrequent or weekend visitors. Complete response from the selected 20 removes nonresponse within that sample; it does not give the other members a selection chance. A complete roster is a possible sampling frame, not evidence that this selection was random.