W8 surveys media public opinion

what are surveys what are they good for

asking questions from random sampling to gather info and statistically analyse

survey types?

descriptive (ppl do and think), analytic(research questions and hypothesis) cross sectional (one point in time), longitudinal (across time)

whats opinion polling used for in politics

predict who will win

measure party support

policy/issue polls

why and what is finding an average of polls good for

as polls can be conducted with different methods and at diff times, an average can make analysing change over time easier

what can consumer market research measure

brand recognition, purchasing preferences and trends, what is liked,

what is most common student market research surveys used

NSS national student survey

single issue surveys

module and program survey

who can academic research be done by

unis , gov, independent institutions and centres, think tanks, survey data sets shared amongst researchers ie the BSA core questions

survey advantages

measurements of things we can’t observe naturally ie due to being able to see a large or very small or specific population ,

convienence, fairly cheap

can reach vulnerable population groups due to flexibility, anonymity, targeted sampling

survey disadvantages

multiple points of error in process

creating a survey is not exact science

deliverate or accidental misrepresentation of results

what is the data types

nominal

dichotimous

ordinal

interval

what is nominal data type

variables with two or more cateories

what is dichotomous data type

two categories

what is ordinal data type

ranked variables e.g strongly agree

what is interval data types

numerical thats continuous , continuous and dicrete type (discrete means finite range)

what are the diff answer structures u could have

single response

multiple response

limitied multiple response

ranked response

scale response

open response

what are diff ways u can administer surveys

self administered

group administered

interview survey

what makes a strong survey design

time series method, identical questions yearly

new questions robustly tested

random probability sampling

multi channel completion

data checks

funded

how to generate reliable responses

make sure knowledge is accessible so people dont guess, do path of least resitance

social desirability, ppl claim to be better than they are , ie the shy tory effect