Download CMN3102 Midterm Exam: Social Science Research Methods and Data Analysis and more Exams Advanced Education in PDF only on Docsity!
CMN3 102 Midterm Exam With
Complete Solution
Human computer - ANSWER one who computes
First digital computer - ANSWER Abacus
Atanasoff-Berry Computer (ABC) - ANSWER Invented in 1937 by Atanasoff to speed up solutions of math problems. There was no patent on it which allowed people to steal his ideas. Went to court in 1973 and recognized John Atanasoff as the inventor of the electronic digital computer which annuled ENIAC patents.
ENIAC - ANSWER Electronic Numerical Integrator and Computer. Credited to Eckert and Mauchly as the first electronic computer but could be traced to Atanasoff because Mauchly borrowed his manuscript.
Charles Cole - ANSWER Patent lawyer that had to prove the ENIAC was based on Atanasoff's computer.
Variables - ANSWER Provide quantitative measurment of concepts. They contain different values.
Nominal Variable - ANSWER A variable that has values that cannot be ranked and cannot be compared as higher or lower. Essentially a category. Ex: Countries.
Ordinal Variables - ANSWER A variable with values we can rank or order along some dimension but cannot find the average (the mean) because the distance between each value is not the same. Ex: Levels of education - HS, Uni, Post-grad.
Interval/ratio variables - ANSWER A variable whose values can be ranked/ordered and have equal distances between adjacent values. Ex: income measured in dollars.
Happiness in countries video - ANSWER - Based on many factors like democracy, economic growth, and personal choice.
- Canadians are happier than americans
- Rank among the top happiest countries in the world
Survey design - ANSWER 1. Research question and/or hypothesis
- Select sample
- Questionnaire must be created using official languages.
- Type of survey must be chosen.
- Pilot testing.
Reserach hypothesis - ANSWER A statement that we want to prove or disprove with the help of research
Research question - ANSWER A question that we want to answer with the help of research.
Population - ANSWER All cases about which one seeks knowledge. Ex: all uOttawa students
Volunteer sample (non-probability sampling) - ANSWER Self-selected participants / respondents. Ex: The Greatest Canadian reality TV contest.
Quota sample - ANSWER Set quotas for specific groups.
Issue with non-probability sampling - ANSWER Difficult to generalize to population. They are not generally representative of the population.
Sampling error - ANSWER Errors of estimation that occur as a result of differences between the characteristics of the sample and those of the population
Non-sampling error - ANSWER Occurs when people lie or make up their responses
Non-response - ANSWER Occurs if people selected for the sample do not supply the required data.
Response rate - ANSWER The percentage of the sample that participates in the study. An effort should be made to determine whether people who are selected for the sample but decline to participate differ in some important way from those who participate
Sample size - ANSWER It is the absolute size of the sample that matters, not the proportion of the population that it comprises. As sample esize increases, sampling error tends to decrease.
Typical sample sizes in national surveys - ANSWER 1000, 1600, 2500
Standardized/structured interviews and questionnaires - ANSWER Used in survey research to produce standardization in the asking of questions and recording of answers. They reduce error due to interviewer variability. Characteristics of the interviewer may influence responses.
Coding - ANSWER transforming measures into numbers. Ex: high confidence in TV is given a code of '4' which is then recorded in a file.
Conducting interviews: basic points. - ANSWER 1. Ask questions exactly as stated.
- Record answers as exactly as possible
- Questions should be asked in the order of the interview schedule
- Can use closed and open-ended questions
- Keep in mind that leading questions can affect responses
Leading questions - ANSWER Asking a bunch of questions in order to get the response you really want (ex: "Yes Prime Minister" video). Leading questions are basically an example of fraud.
Respondent problem - Acquiesence - ANSWER The respondent is just trying to please the researcher. Ex: agreeing with all items just to be cooperative.
Disadvantages of online surveys - ANSWER - high dropout rate (people forget about it)
- Many countries do not have internet access
- Older people less likely to respond
Advantages of face to face surveys - ANSWER - Can be done at home
- Can be done on the street
- More representative than phone or online interviews in many countries
Disadvantages of face to face surveys - ANSWER - Time consuming
- Very costly
- Pandemic made these hard to conduct over the last few years
Advantages of mail surveys - ANSWER - No interviewees needed as it is sent by mail
Disadvantages of mail surveys - ANSWER - Relatively expensive
- Low response rate in many cases
- Many illeterate respondents in various countries
Omnibus surveys - ANSWER Combination of many small surveys.
Census - ANSWER Surveys an entire population (ex: Sandy Hill community). Very expensive to conduct.
Frequency distribution - ANSWER Percentage distribution of the variable of interest. Use tables or charts to summarize the distribution of a variable. Used for all types of variables.
Reliability - ANSWER Achieved if the same results are found each time a particular measurement technique is used on the same subject (assuming that what is being measured has not changed). Ex: if you have the same representative sample, repeated results from a polling survey for election should be very similar.
Replicability - ANSWER Achieved if others are able to repeat all or part of a study and get the same results. Not about measurment, it is about the ability to repeat the same result. Ex: found multiple times by multiple reserachers that people are more succesful after doing the "power pose".
Validity - ANSWER There is integrity to the conclusions
Measurement validity - ANSWER AKA construct validity. Are you measuring what you want to measure (the idea/concept in the question)? Ex: if a reseracher is prejudiced certain subjects may be favoured and some may not tell the truth therefore not being a valid result.
Median - ANSWER Middle value. Only used for ordinal and interval-ratio variables. Can be identified as middle value or value corresponding to 50% of the cases in the cumulative percent frequency distribution.
Mean - ANSWER Average value. Can only be used for interval-ratio variables.
Dependent variable - ANSWER A variable which we are interested in explaining. Assumed to depend on or be affected by the IV. Ex: confidence in TV.
Independent variable - ANSWER A variable which we use to explain the DV. Presumed to affect the DV. Ex: level of education.
Research hypothesis (H1) - ANSWER A statement about the relationship between the IV and the DV that we want to prove or disprove.
Null hypothesis (H0) - ANSWER Opposite of research hypothesis. A statement of "no difference" between IV and DV.
Statistically non-significant relationship between IV and DV = ... - ANSWER null hypothesis is true and the research hypothesis is rejected.
Statistically significant relationship between IV and DV and in expected direction if
specified = ... - ANSWER Null hypothesis is false and research hypothesis is true.
Result is opposite to direction specific in hypothesis = ... - ANSWER research hypothesis is rejected.
Bivariate analysis - ANSWER Determines whether there is a relationship between the DV and IV.
Cross-tabulation - ANSWER Mostly used in bivariate analysis. Most appropriate for nominal and ordinal variables.
Bivariate table of percentages - ANSWER Similar to distribution table. DV is in rows, IV is in columns. Percentage totals are column totals equal to 100%. You compare percentages in the columns to determine if data is statistically significant
Rule of thumb - ANSWER Differences are statistically significant if percentage differences in column percentages are more than 4% in at least one case in cross-tab table. Differences are not statistically significant if percentage differences in column percentages are 4% or less in all cases in cross-tab table.
Measures of association - ANSWER Provide info about the strength of relationship between variables and the direction of association (in the case of ordinal or interval-ratio variables)
Less than 0.001 p value (statistical significance) - ANSWER probability that a tested relationship occurred by chance is less than .001, or 1 in 1000, or 0.1%
Less than 0.01 p value (statistical significance) - ANSWER probability that a tested relationship occurred by chance is less than .01, or 1 in 100, or 1%
Less than 0.05 p value (statistical significance) - ANSWER probability that a tested relationship occurred by chance is less than .05, or 1 in 20, or 5%
Less than 0.10 p value (statistical significance) - ANSWER can be used if N is small
Reasons for ABC patent failure - ANSWER - Iowa state didn't file for patent because of WW
- Attanasoff left for war
- Mauchly was assinged to create a computer for WW2, used Attanasoff's model
- No patent because Attanasoff didn't sign anything before he left
