1What is the first step in the sampling design process?
Sampling design process
Easy
A.Writing the research report
B.Defining the target population
C.Interpreting the collected data
D.Calculating the sample results
Correct Answer: Defining the target population
Explanation:
The researcher must first clearly identify the population about which conclusions will be drawn.
Incorrect! Try again.
2What is a sampling frame?
Sampling design process
Easy
A.A summary of research findings
B.A schedule for data collection
C.A list of population elements
D.A table of statistical results
Correct Answer: A list of population elements
Explanation:
A sampling frame is the list or source from which members of the sample are selected.
Incorrect! Try again.
3What does determining the sample size involve?
Sampling design process
Easy
A.Selecting the final research title
B.Deciding how many units to study
C.Identifying the dependent variable
D.Choosing how results are displayed
Correct Answer: Deciding how many units to study
Explanation:
Sample size refers to the number of individuals, objects, or units included in the sample.
Incorrect! Try again.
4Why is the sampling unit specified in a sampling design?
Sampling design process
Easy
A.To explain the study limitations
B.To calculate the research budget
C.To summarize the collected responses
D.To identify the element being selected
Correct Answer: To identify the element being selected
Explanation:
A sampling unit identifies the basic element, such as a person or household, that may be selected.
Incorrect! Try again.
5Which characteristic means that a sample closely reflects its population?
Characteristics of a good sample
Easy
A.Representativeness
B.Complexity
C.Uniformity
D.Convenience
Correct Answer: Representativeness
Explanation:
A representative sample reflects the important characteristics of the target population.
Incorrect! Try again.
6A good sample should have which type of sampling error?
Characteristics of a good sample
Easy
A.Unknown sampling error
B.Increasing sampling error
C.Maximum sampling error
D.Minimum sampling error
Correct Answer: Minimum sampling error
Explanation:
A good sample minimizes the difference between sample estimates and true population values.
Incorrect! Try again.
7What does an adequate sample size help improve?
Characteristics of a good sample
Easy
A.Length of the questionnaire
B.Number of research topics
C.Style of the final report
D.Reliability of estimates
Correct Answer: Reliability of estimates
Explanation:
An adequate sample size generally produces more stable and reliable estimates of population characteristics.
Incorrect! Try again.
8Which quality makes a sample practical to study with available resources?
Characteristics of a good sample
Easy
A.Complexity
B.Feasibility
C.Uniformity
D.Randomness
Correct Answer: Feasibility
Explanation:
A feasible sample can be studied within the available time, budget, and staffing limits.
Incorrect! Try again.
9What are the two broad types of sampling design?
Types of sampling design
Easy
A.Internal and external sampling
B.Probability and non-probability sampling
C.Primary and secondary sampling
D.Qualitative and quantitative sampling
Correct Answer: Probability and non-probability sampling
Explanation:
Sampling designs are broadly classified as probability sampling and non-probability sampling.
Incorrect! Try again.
10In probability sampling, how are population units selected?
Types of sampling design
Easy
A.By a random selection process
B.By the researcher's preference
C.By ease of accessibility
D.By participant recommendations
Correct Answer: By a random selection process
Explanation:
Probability sampling uses random selection so that units have known chances of inclusion.
Incorrect! Try again.
11Which feature is typical of non-probability sampling?
Types of sampling design
Easy
A.Every unit receives a number
B.Selection chances are unknown
C.Selection chances are always equal
D.Every subgroup is sampled randomly
Correct Answer: Selection chances are unknown
Explanation:
In non-probability sampling, the probability that each population unit will be selected is not known.
Incorrect! Try again.
12Which sampling design is generally more suitable for statistical generalization?
Types of sampling design
Easy
A.Snowball sampling
B.Probability sampling
C.Convenience sampling
D.Judgment sampling
Correct Answer: Probability sampling
Explanation:
Probability sampling supports statistical generalization because selection is based on known probabilities.
Incorrect! Try again.
13Which technique gives every population member an equal chance of selection?
Random sampling techniques
Easy
A.Judgment sampling
B.Simple random sampling
C.Quota sampling
D.Snowball sampling
Correct Answer: Simple random sampling
Explanation:
Simple random sampling gives each member of the population an equal chance of being selected.
Incorrect! Try again.
14Which sampling technique selects every th unit from an ordered list?
Random sampling techniques
Easy
A.Purposive sampling
B.Stratified sampling
C.Cluster sampling
D.Systematic sampling
Correct Answer: Systematic sampling
Explanation:
Systematic sampling selects units at a regular interval, such as every th unit, after a starting point is chosen.
Incorrect! Try again.
15In which technique is the population divided into homogeneous subgroups before sampling?
Random sampling techniques
Easy
A.Convenience sampling
B.Stratified sampling
C.Snowball sampling
D.Systematic sampling
Correct Answer: Stratified sampling
Explanation:
Stratified sampling divides the population into similar subgroups called strata and samples from each stratum.
Incorrect! Try again.
16Which technique randomly selects entire groups, such as schools or villages?
Random sampling techniques
Easy
A.Systematic sampling
B.Quota sampling
C.Judgment sampling
D.Cluster sampling
Correct Answer: Cluster sampling
Explanation:
Cluster sampling divides the population into groups and randomly selects some of those groups.
Incorrect! Try again.
17Which technique selects participants because they are easy to reach?
Non-random sampling techniques
Easy
A.Systematic sampling
B.Cluster sampling
C.Convenience sampling
D.Stratified sampling
Correct Answer: Convenience sampling
Explanation:
Convenience sampling includes participants who are readily available or easily accessible.
Incorrect! Try again.
18Which technique selects participants based on the researcher's expert judgment?
Non-random sampling techniques
Easy
A.Purposive sampling
B.Simple random sampling
C.Cluster sampling
D.Systematic sampling
Correct Answer: Purposive sampling
Explanation:
Purposive sampling uses the researcher's judgment to select participants with relevant characteristics.
Incorrect! Try again.
19Which technique asks existing participants to recommend other participants?
Non-random sampling techniques
Easy
A.Quota sampling
B.Systematic sampling
C.Stratified sampling
D.Snowball sampling
Correct Answer: Snowball sampling
Explanation:
In snowball sampling, current participants help recruit additional participants from their networks.
Incorrect! Try again.
20Which technique selects participants until predetermined category totals are reached?
Non-random sampling techniques
Easy
A.Quota sampling
B.Cluster sampling
C.Simple random sampling
D.Systematic sampling
Correct Answer: Quota sampling
Explanation:
Quota sampling fills predetermined numbers of participants in specified population categories.
Incorrect! Try again.
21A researcher wants to study job satisfaction among 2,400 employees working in six departments. Before choosing a sampling technique, what should the researcher do first?
Sampling design process
Medium
A.Replace employees who refuse to respond
B.Define the target population and sampling unit
C.Calculate the final sampling error
D.Select every tenth employee from payroll
Correct Answer: Define the target population and sampling unit
Explanation:
A sampling design begins by clearly defining the population to be studied and the basic unit that may be selected.
Incorrect! Try again.
22A university survey defines its population as all currently enrolled students but uses an outdated hostel list as its sampling frame. Which problem is most likely?
Sampling design process
Medium
A.Recall error
B.Coding error
C.Coverage error
D.Interviewer bias
Correct Answer: Coverage error
Explanation:
The hostel list excludes students living elsewhere and may include ineligible names, so it does not adequately cover the target population.
Incorrect! Try again.
23A researcher has a fixed budget and requires estimates for both urban and rural households. Which design decision best addresses this requirement?
Sampling design process
Medium
A.Treat residence as a basis for stratification
B.Survey only the least expensive urban area
C.Use one national convenience sample
D.Let interviewers choose accessible households
Correct Answer: Treat residence as a basis for stratification
Explanation:
Stratifying by residence ensures that both urban and rural groups are represented while allowing the sample allocation to reflect budget constraints.
Incorrect! Try again.
24A sample size calculation gives 384 respondents, but a response rate of 80% is expected. Approximately how many people should be contacted?
Sampling design process
Medium
A.768
B.307
C.384
D.480
Correct Answer: 480
Explanation:
The required number to contact is , allowing for the expected 20% nonresponse.
Incorrect! Try again.
25Two samples are drawn from the same population. Sample X closely reflects the population's age and income distribution, while Sample Y mainly includes young, high-income respondents. Which quality does Sample X demonstrate more strongly?
Characteristics of a good sample
Medium
A.Representativeness
B.Homogeneity
C.Convenience
D.Administrative simplicity
Correct Answer: Representativeness
Explanation:
A representative sample resembles the population on characteristics relevant to the study, reducing the risk of biased conclusions.
Incorrect! Try again.
26A researcher repeatedly draws samples using the same design. The resulting estimates vary only slightly around the population value. Which two qualities are mainly demonstrated?
Characteristics of a good sample
Medium
A.Precision and accuracy
B.Simplicity and accessibility
C.Diversity and flexibility
D.Convenience and speed
Correct Answer: Precision and accuracy
Explanation:
Small variation indicates precision, while estimates centered near the true population value indicate accuracy.
Incorrect! Try again.
27Increasing a simple random sample from 100 to 400 observations, with population variability unchanged, will approximately have what effect on the standard error?
Characteristics of a good sample
Medium
A.It will remain unchanged
B.It will be doubled
C.It will be halved
D.It will be quartered
Correct Answer: It will be halved
Explanation:
Standard error is approximately proportional to . Increasing fourfold reduces the standard error by .
Incorrect! Try again.
28A large online poll allows anyone visiting a news website to participate. Why may its large sample still produce poor estimates of public opinion?
Characteristics of a good sample
Medium
A.Public opinion requires a complete census
B.Large samples always increase sampling bias
C.Online responses cannot be analyzed statistically
D.Self-selection can create systematic bias
Correct Answer: Self-selection can create systematic bias
Explanation:
A large sample reduces random error but does not correct selection bias when participation depends on respondents volunteering themselves.
Incorrect! Try again.
29A national study randomly selects districts, then schools within selected districts, and finally students within selected schools. Which sampling design is being used?
Types of sampling design
Medium
A.Quota sampling
B.Judgment sampling
C.Systematic sampling
D.Multistage sampling
Correct Answer: Multistage sampling
Explanation:
Multistage sampling selects units through successive levels, such as districts, schools, and then students.
Incorrect! Try again.
30A city is divided into 80 neighborhoods. Ten neighborhoods are randomly selected, and every household in those neighborhoods is surveyed. Which design is this?
Types of sampling design
Medium
A.Quota sampling
B.Systematic random sampling
C.One-stage cluster sampling
D.Proportionate stratified sampling
Correct Answer: One-stage cluster sampling
Explanation:
The neighborhoods are clusters, and surveying every household in each selected neighborhood makes this one-stage cluster sampling.
Incorrect! Try again.
31Which situation most strongly favors stratified sampling rather than cluster sampling?
Types of sampling design
Medium
A.A complete list exists only for geographic groups
B.Natural groups are miniature versions of the population
C.Subgroups differ internally and each needs representation
D.Travel costs dominate all other research concerns
Correct Answer: Subgroups differ internally and each needs representation
Explanation:
Stratification is useful when meaningful subgroups differ from one another and the researcher needs adequate representation from each subgroup.
Incorrect! Try again.
32A company samples 10% of employees from each department, with the number chosen from each department proportional to its size. What design is used?
Types of sampling design
Medium
A.Disproportionate stratified sampling
B.Single-stage cluster sampling
C.Fixed quota sampling
D.Proportionate stratified sampling
Correct Answer: Proportionate stratified sampling
Explanation:
Departments form strata, and selecting the same percentage from each means sample allocation is proportional to stratum size.
Incorrect! Try again.
33A population list contains 1,200 names, and a systematic sample of 100 is required. After choosing a random start from 1 to 12, which interval should be used?
Random sampling techniques
Medium
A.Every 12th name
B.Every 10th name
C.Every 120th name
D.Every 100th name
Correct Answer: Every 12th name
Explanation:
The systematic interval is , so every 12th name is selected after a random start.
Incorrect! Try again.
34A factory uses systematic sampling on a production list where every 20th item comes from the same machine. The sampling interval is also 20. What is the main risk?
Random sampling techniques
Medium
A.Periodicity may bias the sample
B.Random starts become impossible
C.Each item will have equal representation
D.The sample size will necessarily double
Correct Answer: Periodicity may bias the sample
Explanation:
When the list has a repeating pattern matching the sampling interval, systematic sampling may repeatedly select items from one machine.
Incorrect! Try again.
35In simple random sampling without replacement, 50 students are selected from 500. What is the probability that any particular student is included?
Random sampling techniques
Medium
A.
B.
C.
D.
Correct Answer:
Explanation:
Under simple random sampling, each student's inclusion probability is .
Incorrect! Try again.
36A small occupational group makes up 2% of a population but must be analyzed separately. Which probability approach is most appropriate?
Disproportionate allocation can oversample a small stratum so that enough cases are available for reliable subgroup analysis.
Incorrect! Try again.
37A researcher studying an undocumented migrant community asks each participant to refer other eligible participants. Which method is being used?
Non-random sampling techniques
Medium
A.Systematic sampling
B.Stratified sampling
C.Quota sampling
D.Snowball sampling
Correct Answer: Snowball sampling
Explanation:
Snowball sampling uses referrals from existing participants to reach members of populations that are difficult to identify or access.
Incorrect! Try again.
38Interviewers must recruit 60 women and 40 men at a shopping center, but they may choose any available individuals within each category. Which method is this?
Non-random sampling techniques
Medium
A.Simple random sampling
B.Quota sampling
C.Stratified random sampling
D.Cluster sampling
Correct Answer: Quota sampling
Explanation:
Quota sampling sets target numbers for categories but does not randomly select individuals within those categories.
Incorrect! Try again.
39A researcher deliberately selects experienced emergency physicians because they possess specialized knowledge relevant to the study. Which technique is most appropriate?
Non-random sampling techniques
Medium
A.Systematic sampling
B.Cluster sampling
C.Purposive sampling
D.Convenience sampling
Correct Answer: Purposive sampling
Explanation:
Purposive sampling deliberately selects participants whose characteristics or expertise directly serve the research objective.
Incorrect! Try again.
40A lecturer surveys students who remain in the classroom after an optional seminar because they are easiest to reach. What is the main limitation of this method?
Non-random sampling techniques
Medium
A.The sample requires a complete student sampling frame
B.The method guarantees excessive financial cost
C.The method gives every student an equal probability
D.The sample may differ systematically from other students
Correct Answer: The sample may differ systematically from other students
Explanation:
This is convenience sampling, and readily available students may differ in motivation or interest from those who did not attend or remain.
Incorrect! Try again.
41A national study defines its target population as all adults who usually reside in private households. Its sampling frame is an electoral register that excludes nonregistered adults and retains some people who have moved abroad. Which pair of errors is present before any sampled person is contacted?
Sampling design process
Hard
A.Sampling variance and overcoverage
B.Undercoverage and overcoverage
C.Item nonresponse and measurement error
D.Unit nonresponse and undercoverage
Correct Answer: Undercoverage and overcoverage
Explanation:
Excluded eligible adults create undercoverage, while listed ineligible former residents create overcoverage. Both are frame errors rather than nonresponse or measurement errors.
Incorrect! Try again.
42A researcher must estimate a population mean with margin of error at 95% confidence. A pilot study gives . Ignoring finite population correction and nonresponse, what is the smallest required sample size using with ?
Sampling design process
Hard
A.135
B.139
C.142
D.148
Correct Answer: 139
Explanation:
. The result must be rounded upward, giving .
Incorrect! Try again.
43A probability sample is selected correctly, but only 60% respond. Response propensity depends strongly on age, which is known for every sampled unit, and age also predicts the outcome. Which revision most directly addresses the resulting risk during estimation?
Sampling design process
Hard
A.Increase every respondent's weight by the same factor
B.Report the unweighted mean with a smaller confidence level
C.Replace missing outcomes with the overall respondent mean
D.Adjust weights within age-based response classes
Correct Answer: Adjust weights within age-based response classes
Explanation:
Response-class weighting uses observed age information related to both response and outcome, reducing nonresponse bias under a conditional response assumption.
Incorrect! Try again.
44A firm wants separate estimates for three rare occupational groups as well as an overall workforce estimate. A simple random sample would include too few members of each rare group. Which design-stage decision best meets both objectives?
Sampling design process
Hard
A.Oversample the rare groups and analyze all cases unweighted
B.Create convenience quotas for the rare groups at each worksite and treat the resulting cases as a self-weighting probability sample
C.Sample only the rare groups and impute the remaining workforce
D.Oversample the rare groups and use design weights
Correct Answer: Oversample the rare groups and use design weights
Explanation:
Disproportionate stratified sampling supplies adequate subgroup cases, while inverse-probability design weights restore valid overall population estimation.
Incorrect! Try again.
45Two unbiased designs estimate the same population mean. Design A has variance and expected cost ; Design B has variance and expected cost . If quality is judged by minimizing , which design is more efficient?
Characteristics of a good sample
Hard
A.Design B, because its criterion value is
B.Design A, because its standard error is
C.Design A, because its criterion value is
D.Design B, because lower variance always dominates cost
Correct Answer: Design A, because its criterion value is
Explanation:
The criterion is for A and for B. The smaller product indicates greater cost-adjusted efficiency.
Incorrect! Try again.
46A very large online opt-in sample closely matches population margins for age, sex, and region after weighting. Which conclusion is most defensible?
Characteristics of a good sample
Hard
A.Weighting converts every volunteer sample into a probability sample with known inclusion probabilities and therefore supports exact randomization-based inference
C.Bias may remain from unmeasured selection-related variables
D.Large size guarantees negligible selection bias
Correct Answer: Bias may remain from unmeasured selection-related variables
Explanation:
Calibration can correct imbalance on observed variables, but it cannot guarantee correction for unobserved factors associated with volunteering and the study outcome.
Incorrect! Try again.
47A cluster sample has respondents per cluster and an intracluster correlation of . Under equal cluster sizes, what is the approximate design effect relative to simple random sampling?
Characteristics of a good sample
Hard
A.
B.
C.
D.
Correct Answer:
Explanation:
For equal-sized clusters, .
Incorrect! Try again.
48A sample estimate has a very small reported standard error, but the frame omits 20% of the target population whose outcome values are systematically higher. Which assessment is correct?
Characteristics of a good sample
Hard
A.The estimate is representative because its standard error is small
B.The estimate is unbiased but statistically inefficient
C.The estimate is precise but potentially biased
D.The estimate becomes unbiased once the nominal sample size exceeds the omitted population count
Correct Answer: The estimate is precise but potentially biased
Explanation:
A small standard error reflects low sampling variability under the implemented design; it does not remove systematic undercoverage bias.
Incorrect! Try again.
49A researcher randomly selects 30 hospitals, then independently draws a simple random sample of nurses within each selected hospital. How should this design be classified?
Types of sampling design
Hard
A.Stratified random sampling
B.One-stage cluster sampling
C.Systematic element sampling
D.Two-stage cluster sampling
Correct Answer: Two-stage cluster sampling
Explanation:
Hospitals are sampled as primary sampling units, followed by sampling nurses as secondary units, making the design two-stage cluster sampling.
Incorrect! Try again.
50A population is divided into mutually exclusive income bands, and an independent random sample is taken from every band. Which feature distinguishes this design from cluster sampling?
Types of sampling design
Hard
A.Only selected strata contribute sampled units
B.Every stratum contributes sampled units
C.Income bands must have equal population sizes
D.Every element receives the same final inclusion probability regardless of the stratum allocation selected by the researcher
Correct Answer: Every stratum contributes sampled units
Explanation:
In stratified sampling, units are sampled from all strata. In standard cluster sampling, only selected clusters contribute elements.
Incorrect! Try again.
51A survey selects villages with probability proportional to population size and then selects exactly 20 households by simple random sampling within each chosen village. Ignoring repeated village selection, when is the household sample approximately self-weighting?
Types of sampling design
Hard
A.When every village has the same geographic area
B.When village selection probability is proportional to household count
C.When the number of selected villages is proportional to the observed response rate in each village
D.When household outcomes have equal within-village variance
Correct Answer: When village selection probability is proportional to household count
Explanation:
The first-stage probability is proportional to village size, while the second-stage probability is inversely proportional to that size. Their product is then approximately constant.
Incorrect! Try again.
52Which design is best described as a probability design even though sampled units can have unequal inclusion probabilities?
Types of sampling design
Hard
A.Snowball sampling that continues until no participant supplies a new referral and all observed network components appear saturated
B.Probability-proportional-to-size sampling with known measures of size
C.Purposive sampling of cases judged to be maximally informative
D.Quota sampling using interviewer-selected cases within demographic cells
Correct Answer: Probability-proportional-to-size sampling with known measures of size
Explanation:
Probability sampling requires known, nonzero selection probabilities, not equal probabilities. PPS sampling deliberately assigns unequal but known probabilities.
Incorrect! Try again.
53A population has two strata with and , where is per-unit cost. Under optimum allocation with unequal costs, . For total , what allocation is optimal?
Random sampling techniques
Hard
A.
B.
C.
D.
Correct Answer:
Explanation:
The allocation measures are and , a ratio of . Thus and .
Incorrect! Try again.
54A systematic sample of every 10th unit is drawn after one random start from a production list arranged in a repeating cycle of exactly 10 machine positions. What is the central design risk?
Random sampling techniques
Hard
A.Every unit obtains an inclusion probability greater than
B.The estimator is automatically unbiased because a random start eliminates any effect of ordering or periodicity in the frame
C.The sample may repeatedly select one machine position
D.The sampling interval becomes random after the first selection
Correct Answer: The sample may repeatedly select one machine position
Explanation:
When the frame's period matches the sampling interval, one random start can lock the sample onto a single position in each cycle, producing severe bias.
Incorrect! Try again.
55In unequal-probability sampling without replacement, unit has inclusion probability . Which estimator of the population total is design-unbiased under the sampling design?
Random sampling techniques
Hard
A.
B.
C.
D.
Correct Answer:
Explanation:
The Horvitz-Thompson estimator weights each observed value by the inverse of its inclusion probability and is design-unbiased for the population total.
Incorrect! Try again.
56From a finite population of , an SRS without replacement of produces an estimated standard error of when the finite population correction is ignored. What is the corrected standard error using ?
Random sampling techniques
Hard
A.
B.
C.
D.
Correct Answer:
Explanation:
The correction is , so the corrected standard error is .
Incorrect! Try again.
57Interviewers must fill age-by-sex quotas but may approach anyone they prefer within each quota cell. Why does this remain non-random sampling?
Non-random sampling techniques
Hard
A.Within-cell inclusion probabilities are unknown
B.The final sample cannot contain the same number of respondents in every age-by-sex group unless replacement sampling is explicitly permitted
C.Quota cells are mutually exclusive
D.Population proportions are used to set targets
Correct Answer: Within-cell inclusion probabilities are unknown
Explanation:
Matching quota totals does not make interviewer selection probabilistic. Without known within-cell selection probabilities, design-based inclusion weights cannot be derived.
Incorrect! Try again.
58A study of an undocumented and socially connected population recruits initial participants purposively and asks each participant to refer others. Which limitation most directly threatens estimation of population prevalence?
Non-random sampling techniques
Hard
A.Sampling variance must be zero because participants recruit acquaintances with similar characteristics
B.The method requires a complete conventional sampling frame
C.Recruitment depends on network ties and seed selection
D.Every referral necessarily has the same selection probability
Correct Answer: Recruitment depends on network ties and seed selection
Explanation:
Snowball recruitment overrepresents well-connected networks and can remain strongly dependent on the initial seeds, making population prevalence difficult to estimate.
Incorrect! Try again.
59A qualitative researcher deliberately chooses cases that sharply differ on organizational size, ownership, and region to examine whether a proposed mechanism appears across diverse contexts. Which technique is being used?
Maximum-variation sampling intentionally selects diverse cases to investigate common patterns and contextual differences; it does not rely on random selection.
Incorrect! Try again.
60An opt-in web poll reports that 72% support a policy, with a conventional SRS margin of error of . Which interpretation is methodologically strongest?
Non-random sampling techniques
Hard
A.The margin of error omits unknown self-selection bias
B.The stated margin fully captures coverage and nonresponse errors
C.The margin remains valid after weighting because demographic calibration reconstructs all unknown inclusion probabilities for every member of the target population
D.The estimate is design-unbiased because the sample is large
Correct Answer: The margin of error omits unknown self-selection bias
Explanation:
A conventional SRS margin measures random sampling error under a probability model. It does not quantify potentially substantial opt-in self-selection and coverage biases.
Incorrect! Try again.
Did this save you a night before the exam?
LPU Notes is free, and it stays free. Ads cover part of the server bill.
The rest comes out of a student's own pocket: the domain, the storage,
and keeping the site up through the weeks everyone needs it at once.
The payment button didn't load. An ad blocker or a filtered network is the usual reason.
to try again.
Nothing here is ever locked, and nothing unlocks. Chip in only if it was worth it.
What it pays for →