UGC NET Questions (Paper – 1)

Non-sampling errors are all sources of survey inaccuracy that do not stem from the act of sampling but from the way data are collected, recorded or processed. They include interviewer bias, faulty questionnaire design, respondent misunderstanding and coding or tabulation mistakes. Such errors can be large and may not decrease simply by increasing sample size. Therefore, errors arising from inaccurate responses or defective tools are correctly termed non-sampling errors.

Option A:
Non-sampling errors encompass diverse problems such as ambiguous questions that confuse respondents, poorly trained interviewers and data entry errors. These issues affect the quality of information gathered from each selected unit and can seriously distort findings. Since the stem emphasises inaccurate responses and processing mistakes, this option is appropriate.

Option B:
Random error refers to unpredictable fluctuations that occur by chance and may be part of both sampling and measurement processes; it is not restricted to the specific systematic problems listed.

Option C:
Proportional error is not a standard category in survey methodology and does not describe the cluster of issues like tool defects and processing mistakes.

Option D:
Experimental error is associated with variability in experimental outcomes and does not specifically refer to survey response or instrument problems, so it is not correct here.

Reliability focuses on the stability and consistency of measurement. If a test is administered to the same group on two occasions without major changes in conditions, similar scores should be obtained. High reliability means that random error is low and observed scores largely reflect true differences among individuals. This property is fundamental for trustworthy measurement.

Option A:
Option A describes validity, which concerns whether the test measures the intended construct. Although related, validity and reliability are conceptually distinct.

Option B:
Option B is correct because it emphasizes consistency of scores across time or equivalent forms. Techniques such as test–retest and internal consistency estimates are used to evaluate this aspect.

Option C:
Option C refers to criterion related validity, particularly predictive validity, which is about forecasting future outcomes, not about consistency of scores.

Option D:
Option D mentions face validity, which is about appearance of relevance but does not guarantee consistent measurement.

Reliability concerns the stability and consistency of measurement. A reliable test yields similar results under consistent conditions when administered repeatedly to the same group. High reliability means that measurement error is low and observed scores mainly reflect true differences among individuals. It is an important psychometric property but distinct from validity.

Option A:
Option A actually defines validity, which is about whether the instrument measures the intended construct. While validity is crucial, it is not the same as reliability. A tool can be reliable but still not valid.

Option B:
Option B correctly explains reliability as consistency of scores over time or across equivalent forms. Procedures such as test–retest, split half and inter rater reliability are used to estimate this property. When reliability is high, researchers can trust that variations in scores are not primarily due to random error.

Option C:
Option C claims that reliability is about one time administration, which has no conceptual basis. Reliability is specifically concerned with repeated or alternative measurements.

Option D:
Option D suggests that a reliable test always yields high scores, which is unrelated to the concept. Scores may be high or low depending on difficulty; reliability is about consistency, not level.

Statements A, B, C and D distinguish sampling error from non-sampling error, and F highlights a way to reduce some non-sampling problems. A is true because sampling error results from studying only a subset, while B correctly notes that poor instruments create non-sampling error. C is accurate in classifying non-response bias as a non-sampling error, and D correctly observes that larger samples usually reduce sampling error. F is also true since properly trained field staff can lower interviewer and recording errors, whereas E is false because random sampling addresses sampling error, not all forms of non-sampling error.

Option A:
Option A ignores statement F, which draws attention to practical steps for reducing non-sampling error through training. While A, B, C and D are correct, the omission of F makes the description less complete in terms of remedies. Thus, this option is not fully correct.

Option B:
Option B adds statement E, falsely claiming that non-sampling errors can be entirely removed by random sampling. Random sampling influences only sampling error; measurement, non-response and recording errors can still occur. Because it includes a false statement, this option is invalid.

Option C:
Option C is correct because it lists A, B, C, D and F, capturing both definitions and some methods for error reduction. It properly excludes E, which overstates the power of random sampling. Therefore, this combination contains all and only the true statements.

Option D:
Option D omits statement A and includes E, thereby losing the basic definition of sampling error and adding an incorrect assertion. This combination confuses the distinction between error types and cannot be accepted.

Statements A, B, C and D identify key error sources in surveys, and F points to one mitigation strategy. A is true because coverage error arises when parts of the population are missing from the frame, and B correctly defines non-response error. C is accurate in linking measurement error to poor wording or misunderstanding, while D recognises that different modes of administration can affect responses. F is also true since repeated contacts and follow-ups can reduce non-response, whereas E is false because social desirability bias can distort answers and undermine validity.

Option A:
Option A omits statement F, thereby failing to recognise a practical method for reducing non-response error. Although A, B, C and D are correct, the lack of F makes the option less complete in terms of both diagnosis and remedy.

Option B:
Option B leaves out statement A, ignoring the important concept of coverage error, which can prevent some population members from being sampled at all. Without A, the picture of survey validity threats is incomplete, so this option is not acceptable.

Option C:
Option C excludes statement B, neglecting the impact of non-response error on survey conclusions. Even though it includes A, C, D and F, missing B means the combination does not cover all the major threats listed.

Option D:
Option D is correct because it includes A, B, C, D and F, summarising coverage, non-response, measurement and mode effects as well as follow-up strategies. It excludes E, which wrongly claims that social desirability bias has no effect on validity. Therefore, this combination contains all and only the true statements.

Statements A, B, C and E accurately identify sources of measurement error and strategies to reduce it. Poor wording and careless responding can both distort data, while standardised procedures and training investigators help make measurements more consistent. Statement D is false because measurement error directly undermines reliability by introducing random or systematic noise into scores. Therefore, the correct combination includes A, B, C and E but not D.

Option A:
Option A is correct because it includes the four statements that together show how errors arise and how they can be mitigated through design and training. It recognises that reliability depends on minimising such errors.

Option B:
Option B is incomplete as it excludes E, thereby neglecting the role of training in reducing interviewer or observer-related errors. Without E, the focus is limited to wording, respondent behaviour and standardisation but not human skill development.

Option C:
Option C is incorrect because it incorporates D, the claim that measurement error does not affect reliability. This directly contradicts the concept of reliability, which is weakened when error variance is high.

Option D:
Option D is wrong because it retains D and omits B, failing to highlight the impact of respondent carelessness on measurement quality while accepting a false statement about reliability. This leads to an inaccurate understanding of error sources and consequences.