Question 1

A survey question that asks respondents the specific number of times they perform a behavior on a given day uses the ________ level of measurement.

Accepted Answer

The specific number of times a behavior is performed represents a quantitative variable that is measured on a scale where there is a meaningful zero point (i.e. the absence of the behavior). This is known as the ratio level of measurement, which is the highest level of measurement as it allows for all mathematical operations to be applied to the data.

Question 2

The use of parametric statistics such as the t-test or analysis of variances (ANOVA) normally requires that the data be at the ________ level of measurement.

Accepted Answer

Parametric statistics such as the t-test or ANOVA require data to be at the interval or ratio level of measurement. These levels of measurement involve numerical data with equal intervals between values or a true zero point, which is necessary for conducting accurate statistical analyses. Nominal data, on the other hand, are categorical and cannot be numerically measured, and therefore are not appropriate for parametric statistics.

Question 3

Level of measurement can be an issue for the ecological validity of research because

Accepted Answer

If a researcher uses a level of measurement that does not match the level of measurement of the variable in natural settings, the findings may not be applicable or generalizable to real-world situations. It is important for the level of measurement chosen by the researcher to match the level of measurement of the variable in order to maintain ecological validity.

Question 4

As level of measurement increases, the ________ of the measure also increases.

Accepted Answer

As the level of measurement increases (from nominal to ordinal to interval to ratio), the amount of information that can be obtained from the data also increases. This is because higher levels of measurement provide more detailed and precise information.

Question 5

Which of the following considerations should influence the choice of the level of measurement for a variable?

Accepted Answer

All of the above considerations should influence the choice of the level of measurement for a variable. The choice of level of measurement will impact the statistical tests that can be used, as some tests are only appropriate for certain levels of measurement. The desired level of ecological validity of the measure should also be taken into account as it may impact how well the measure reflects real-life situations. Finally, the desired level of information content of the measure is also important as it can impact the accuracy and precision of the data collected. Therefore, all of these considerations should be weighed when choosing the appropriate level of measurement for a variable.

Question 6

Statements such as "Mean scores of participants in the experimental condition were twice as high as those of participants in the control condition" should only be made when the dependent measure is based on ________ level scale.

Accepted Answer

Statements about one score being "twice as high" as another are meaningful only when the scale of measurement is a ratio scale, because ratio scales have a true zero point, allowing for meaningful comparisons of absolute magnitudes.

Question 7

For comparative rating scales to provide reliable and valid results,

Accepted Answer

Both conditions regarding respondent familiarity with stimuli and understanding of the characteristics being rated are important for reliable and valid results. However, the characteristics being rated do not necessarily need to be unidimensional.

Question 8

Comparative rating scales generally provide data at the ________ level of measurement.

Accepted Answer

Comparative rating scales rank items in order of preference or importance, making them an ordinal level of measurement. Nominal level is used for categorical data with no numerical value or order, interval level involves equal intervals between values with an arbitrary zero point, and ratio level involves equal intervals between values with a true zero point.

Question 9

Itemized rating scales are similar to ________ questions.

Accepted Answer

Itemized rating scales are similar to multiple-choice questions because they both provide respondents with a set of predefined options from which they can choose their answer.

Question 10

Itemized rating scales generally provide data at the ________ level of measurement.

Accepted Answer

The answer of Itemized rating scales generally provide data at...

Question 11

Which of the following statements about itemized rating scales is FALSE?

Accepted Answer

Itemized rating scales are not the same as numerical rating scales. Numerical rating scales use numbers to represent the response options, while itemized rating scales use words or phrases.

Question 12

Graphic rating scales

Accepted Answer

Graphic rating scales can be especially useful when collecting data from children because they are often more engaging and easier to understand than other types of rating scales. However, graphic rating scales can still provide numerical data and can be reliable and valid when designed and administered properly. The difficulty for poor readers can be mitigated by using pictures or symbols instead of relying on written instructions.

Question 13

In numerical rating scales, anchors are

Accepted Answer

Anchors in numerical rating scales refer to the verbal labels given to the numbers on the scale. They provide meaning to the numeric values and help respondents interpret the scale and provide accurate responses.

Question 14

The term sensitivity of measurement refers to

Accepted Answer

The term sensitivity of measurement refers to the ability of a measure to detect small differences in the level of a variable. A highly sensitive measure can detect even small changes in the variable being measured, while a less sensitive measure may only be able to detect larger changes. The number of scale points on a rating scale (A) is important in terms of the precision of measurement, but not sensitivity. Highly personal issues (C) relate to the sensitivity of the topic being measured, not the sensitivity of the measurement itself. The ability of a measure to elicit accurate responses (D) is related to validity, but not sensitivity.

Question 15

Which of the following statements about numerical rating scales is TRUE?

Accepted Answer

Research suggests that people prefer numerical rating scales that have between 5 and 9 points, as this provides enough options for discrimination without overwhelming the respondent. Sensitivity of measurement is still an issue with numerical rating scales, and it is better to have anchors at intermediate points on the scale to improve the clarity of the scale. Therefore, options A, B, and D are incorrect.

Question 16

Which of the following statements about item response formats is TRUE?

Accepted Answer

Open-ended questions allow for more detailed and specific responses, whereas closed-ended questions restrict the possible responses to a limited set of options. Therefore, if you want respondents to provide specific information, open-ended questions are probably better. Choice B is false; studies have found high correlations between ratings made using different numbers of response options. Choice C is false; having too many response categories can lead to confusion and reduced reliability. Choice D is false; ranking can be more efficient than paired comparison when there is a large number of stimuli to rate.

Question 17

A(n) ________ is a measure composed of two or more items designed to assess the same variable.

Accepted Answer

A multi-item scale is composed of two or more items designed to assess the same variable. Comparative rating scales allow you to compare one item with another, itemized rating scales asks respondents to provide ratings on each attribute (or item) separately, and multiple-response questions allow respondents to choose more than one option from a list.

Question 18

Which of the following statements about multi-item scales is FALSE?

Accepted Answer

The highest level of measurement provided by multi-item scales is not ordinal, but rather interval or ratio. This is because multi-item scales often use Likert-type response options, which have equal intervals between response options and can therefore be treated as interval or ratio data.

Question 19

A survey question takes the form of "I like vanilla ice cream. (circle a number)" with the response options ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). On which type of multi-item scale is this item most likely to be found?

Accepted Answer

This item is most likely to be found on a Likert scale since it is asking for a degree of agreement or disagreement. Likert scales typically have response options that range from strongly agree to strongly disagree.

Question 20

A survey question takes the form of "Vanilla ice cream is (circle a number)" with a response format of "Good 1 2 3 4 5 6 7 Bad". On which type of multi-item scale is this item most likely to be found?

Accepted Answer

This item is most likely to be found on a Semantic differential scale because it asks respondents to rate vanilla ice cream on a continuum between two opposing feelings (good and bad), rather than asking them to agree or disagree with a statement (Likert), rank a set of items (Thurstone), or indicate agreement with increasingly extreme statements (Guttman).

Quiz 15: Survey Research