Sources of data, reliability and validity, designing questionnaires and data collection sheets, open and closed questions, leading questions, pilots, and cleaning data before processing.

What this dot point is asking

Edexcel codes 1d.01 to 1d.06 require you to know the sources data can come from, the meaning of reliability and validity, how to design questionnaires and data collection sheets, the features of good questions (open and closed, avoiding leading questions, time factors, interview technique), the role of pilots and pre-tests, and why and how data is cleaned before processing. Rewriting a faulty question or set of response boxes, and spotting an error to clean, are classic exam tasks.

Sources of data and collection sheets

Edexcel lists several sources of data (code 1d.01): experimental (laboratory, field or natural), simulation (often using random numbers), questionnaires, observation, reference sources, census and sampling. You should be able to design a suitable data collection sheet (often a tally chart with clear categories) to record the data efficiently. Sources of secondary data must be acknowledged.

Reliability and validity

A bathroom scale that reads differently each time you step on it is unreliable; a survey of "fitness" that only asks about gym visits may be reliable but not valid, because it misses other forms of exercise. Edexcel expects you to comment on both when judging a data collection method.

Designing questionnaire questions

Good questions share clear features:

• Closed questions offer fixed response options, which are quick to analyse; open questions allow any answer, giving richer detail but harder analysis.
• Response boxes must not overlap and must cover every possible value (exhaustive), and should include units and a time frame.
• Questions must be neutral, never leading or emotive.

For example, "How many hours of TV do you watch per day?" needs non-overlapping boxes such as "$0$ to $1$", "more than $1$ up to $2$", "more than $2$", with a stated time frame ("per day"). Boxes such as "$0$ to $1$" and "$1$ to $2$" are faulty because $1$ falls in both.

Leading questions and bias

A leading question steers the respondent towards a particular answer, for example "Don't you agree our excellent service is the best?" Emotive words ("excellent", "delicious") and built-in assumptions bias the results. Edexcel also expects awareness of other planning factors: avoiding biased sources, time factors, and choosing an appropriate interview technique (face to face, phone, online), each with its own advantages and disadvantages.

Pilots and pre-tests

A pilot survey (for questionnaires) or pre-test (for experiments) is a small trial run before the main study. It checks that the questions are clear, the response boxes work, the sampling is practical and the data can be analysed, so problems are found and fixed cheaply before committing to the full investigation.

Cleaning data

Before processing, raw data often needs cleaning (codes 1d.05 to 1d.06). Typical issues are missing data, incomplete responses, values in the wrong format, extraneous symbols (especially in spreadsheets), and outliers or anomalies. You decide whether an outlier is a genuine unusual value or a recording error, and you correct or remove erroneous entries so later calculations are not distorted.