f764 geography skills

?
  • Created by: charlie
  • Created on: 17-06-15 10:54
hypothesis
null (H0) testing for no relationship / alternate (H1) relationship in sample does accurately reflect total statistical population
1 of 41
primary data
unprocessed / not anaylsed / not interpreted / geographer in direct control
2 of 41
secondary data
published documents that have been analysed / interpreted / processed census data / textbooks / research papers / published maps
3 of 41
quality of data
Nominal (lowest quality) 1 or 0 - mode + chi squared / Ordinal (medium quality) ranked - spearmans + median / Inteval (high quality) discrete scale + magnitude - pearsons + mean + SD
4 of 41
GIS Adv
AEGIS 3 with piot study / large amount + variety of data / change scale / dynamic so can cope with change / spatial data to show patterns + trends
5 of 41
Maps (adv + disadv)
plan view with scale + direction / shows what exists visually spatial / flexible scales / can be dated / can be overly detailed
6 of 41
Photos (adv + disadv)
many different views (aerial, ground level, satellite) / captured at specific time / shows variety of data / only at one specific point /
7 of 41
risk assessment 4 stages
1. identify risk 2. evaluate severity 3. control measures to reduce risk 4. re-evaluate severity
8 of 41
accuracy (description + errors)
level at which data is exact and free from error - potential errors include measurement error (repeats) operator error (ind.climate.equip)
9 of 41
reliability (description + errors)
extent to which the sample data reflects the greater total statistical population - potential errors include sampling error (inappropriate frame + too small size)
10 of 41
pilot study (description + adv/disadv)
small scale, preliminary to eval success / weaknesses + improvements of study before full scale is performed - ADV - is time achievable / location appropriate / effectiveness / risk / equip . DISADV - takes time
11 of 41
random sampling
'random number tables to select sample point - every individual having equal chance' / adv - stat. tests + not bias / disadv - not represent total SP (miss points) + long time
12 of 41
Stratified sampling
'underlying patterns in total SP taken into account + proportionally sampled' / adv - no points missing / disadv - possibly bias (false info used), possibly no stat tests, needs info, groups under/over represented
13 of 41
Systematic sampling
'points are selected at regular intervals can be spatial + non spatial' / adv - easy, quick, even coverage / disadv - bias, possibly no stat, interval coincide with data pattern / location, missed variations, regularities not typical
14 of 41
Pragmatic sampling
'sampling is taken where access is available and changes are observed' / adv - realistic + risk assessment based / disadv - bias + possibly not stat test
15 of 41
Units of sampling
point / grid (quadrat) / line (transect) / belt (wider transect)
16 of 41
spatial - choloropleth map
'ratio values are expressed as densities/% with use of areas' / method - standardise values - classes - colour shading - key,title,direction,scale / adv - easy + visual / disadv - concealed in whole area + abrupt at boundary + large dominate
17 of 41
spatial - dot map
'ratio values + use of area' / find base map - decide on dot value + size - key, title, scale, direct / adv - info in areas, visual + spatial, stat analysis / disadv - dots merge
18 of 41
spatial - proportional symbols
' ratio data in specific location' / base map - symbol + scale - size of symbol - key,title,scale,direction / adv - visual + spatial, data recovered, stat tests / disadv - difficult to produce, scale + symbols misread
19 of 41
spatial - isoline / isopleth map
'ration data with points' / base map - number + value of isolines - draw by interpolation (between know values) - number isolines - KTSD / adv - spatial patterns, gradual change, guess points / disadv - assume change constant, need lots of points
20 of 41
spatial - flow line
'ratio data use lines (routed or non routed) to show flow paths' / suitable scale - classes + scale of line (width) - KTSD / adv - visual, quick + easy, shows different variables / disadv - generalised, no stat test, scale carefully chose width
21 of 41
non spatial - table
nominal (names) , interval (%), can use freq distribution for histogram / adv - backbone, clear grouping, stat tests / disadv - no visual
22 of 41
non spatial - bar chart
interval (discrete) width always constant + length varies / adv - visual with temporal + spatial, stat tests, comparable / disadv - only non continuous
23 of 41
stacked/divided bar chart
quantity divided into components of absolute values or proportions / adv - easy to draw, visual, stat tests / disadv - too many divisions confuse
24 of 41
pie charts
% data in segments of whole sample / adv - visual, easily comparable, stat tests with given % / disadv - don't always give %, no temporal change, can confuse, order of division standardised
25 of 41
rose/star charts
show direction and length.width of bars show frequency / adv - shows direction + spatial changes, visual pattern, if accurate + scale can recover stat / disadv - hard to produce width/length scale, not always exact direction
26 of 41
line graph
variations absolute/% X independent and Y dependent / adv - shows trends + anomalies / disadv - only continuos
27 of 41
scatter graph
overall trend (not just line) + simple to show relationship / adv - correlation suggested, anomalies, stat tests / disadv - few points skew data, too many confuse, only 2 variables
28 of 41
triangular graph
3 variables add up to 100% each side of triangle, spatial + temporal / adv - compare 3 variables / disadv - only%, difficult to read + construct
29 of 41
Lorenz curve
shows inequalities (deviation) in distribution of phenomena, lorenz = actual distribution, more concave = greater inequality / adv - visual, multiple distributions, Gini coeff. = degree of inequality / disadv - time for c.freq, no exact quant. value
30 of 41
kite diagram
how phenomena's occurrence changes over distance (bradshaw's) / adv - visual, stat tests, relative phenomena compared / disadv - % change + no absolute values
31 of 41
sketch map
rough map that shows basic info - add essential info, KTSD / adv - visual + easy to interpret / disadv - needs quality TKSD to be effective
32 of 41
mean / median / mode
average of all values - included anomalous data / middle value / most common value
33 of 41
standard deviation
measures central tendency around mean
34 of 41
kurtosis
talled / steeper curve = lower SD compared to normal distribution
35 of 41
distribution
normal (bell shaped curve) / +ve skew = mean greater than mode (upwards hill) / -ve skew = mode greater than mean (downwards hill)
36 of 41
Spearman's rank correlation coefficient
used to show the how strong relationship between two ranked variables / adv - easy + simple, not normal distributed / disadv - poor as no magnitude, no explanation for causes
37 of 41
spearman's rank significance testing
use degree of freedom (classes - 2) and standard significance 0.05 (95%) / adv - are results representative of total SP, whether they are significant + to what degree, reduces thoughts of bias + chance, 95% sure
38 of 41
Pearson's product moment correlation coefficient
show relationship / adv - statistically better, more specific, uses magnitude / disadv - only normally distributed data, hard to construct, no reason given
39 of 41
Chi squared
differences between comparable sets of data / null hypothesis put forward (H0) no correlation / significance test for deviation due to chance / > critical value can reject H0
40 of 41
nearest neighbour analysis
statistically describe distribution of spatial data / 'central place theory' / CLUSTERED (H1) = 0 RANDOM (H0) =1 REGULAR (H1) =2.15 / disadv - changes with measurements, as crow flies? or along roads, services affect
41 of 41

Other cards in this set

Card 2

Front

unprocessed / not anaylsed / not interpreted / geographer in direct control

Back

primary data

Card 3

Front

published documents that have been analysed / interpreted / processed census data / textbooks / research papers / published maps

Back

Preview of the back of card 3

Card 4

Front

Nominal (lowest quality) 1 or 0 - mode + chi squared / Ordinal (medium quality) ranked - spearmans + median / Inteval (high quality) discrete scale + magnitude - pearsons + mean + SD

Back

Preview of the back of card 4

Card 5

Front

AEGIS 3 with piot study / large amount + variety of data / change scale / dynamic so can cope with change / spatial data to show patterns + trends

Back

Preview of the back of card 5
View more cards

Comments

No comments have yet been made

Similar Geography resources:

See all Geography resources »See all skills resources »