Geographical Investigation (GEL Topic 3)

Geographical Investigation Paper 1, 20 marks always gotto give evidence and reasons for every thing u give!! Sampling Methods Probability Sampling Non-probability Sampling Simple Random sampling assign numbers to people then use random number generator Convenience sampling select samples because they are convenient Stratified Random sampling select a sample that has proportional makeup to population, then use random sampling Quota sampling select a sample that has proportional makeup to population, then use convenience sampling removes bias more representative sample used when it’s unnecessary or impractical to select a representative sample can be conducted quickly biased selection of samples unlikely to be representative bonus: systemic random sampling divide population by desired sample number number to get n, interview every nth person Graphs to represent data: - Pie chart - Title, legend, percentage - largest to smallest, clockwise order - Bar graph - Title, Label both axes - Comparative Bar graph - Title, Label both axes, Legend for the comparison (shaded/not) - Histogram - no gap in between - Line graph - Scattergraph

Ways to collect data (field study) ● Questionnaire survey ● Interviews ● Secondary Data ● Field Observation - Mind map, Annotated photographs ○ compare photographs with secondary data? (to show imoact’ ● Word Cloud ● Pedestrian counts ● Land use analysis Accuracy Reliability Validity The precision of measurements How close the data collected is to truth values The ‘repeatability’ of data Consistency of the data Does the data and approach capture the answer to the hypothesis? Does it ‘link’? The extent the data measures what was intended to be tested To improve: ● choose appropriate sites to conduct investigation ● choose appropriate timings of the day ● improve the quality of questions asked To improve: ● collect from a bigger sample size (more representative) ● conduct at multiple sites ● conduct at different times of the day - making sure there are no other variables To improve: ● improving both accuracy and reliability “Evaluate reliability” - can talk about factors that make their results lean one way too much, must justify every factor - must include context!! eg: students should collect data over multiple days, rather than just… “State 3 set of question and response options” - must include instructions! (rank these options from…, choose one of the following etc) “How far does the data support this conclusion” / “Is the conclusion valid?” - must include data!! numbers close ended qns can also be likert scale, ranking

“Measure of central tendency” Mean (average of everything) - includes every value in the data set - can be skewed by outliers Median (middle value) - less affected by outliers - not as sensitive in showing the central location in data set Mode (most frequent value) -