Research Methods in Psychology-II. Week 2. Survey PDF

Research Methods in Psychology-II Özlem Ertan-Kaya, PhD [email protected]  Why Do We Use Surveys? ▪ Surveys are used to get information from people about: ▪ Demographic information (age, gender, income, marital status, etc.), ▪ Feelings, thoughts, attitude...

Research Methods in Psychology-II Özlem Ertan-Kaya, PhD [email protected]  Why Do We Use Surveys? ▪ Surveys are used to get information from people about: ▪ Demographic information (age, gender, income, marital status, etc.), ▪ Feelings, thoughts, attitudes, and beliefs about a particular subject, ▪ Past behaviours, ▪ Future plans « If you want to know what people are doing, OBSERVE; If you want to know what they think, ASK! »  Characteristics of Surveys ▪ Surveys can be specific or more global in their goals: ✓ Giving surveys to university faculty members for housing construction → Local ✓ Using surveys to examine the relationship between happiness and wealth → Global ▪ Low cost ▪ Easy to use. ▪ It is a descriptive or correlational research method. ✓ Knowledge and attitude toward menopause phenomenon among women aged 40–45 years Descriptive ✓ Relationship between psychosocial factors and health behaviours for women experiencing menopause Correlational ▪ It can be deceptive → response rate bias or the questionnare may raise suspicion ▪ No draw cause-effect conclusion Survey Methods Response rate bias occurs when there are a few responses to the survey. Such surveys are less likely to be answered. This raises the question of whether the sample is representative. In addition, if the non- responders have a certain common feature, this makes the research biased. For example, we will evaluate the attitudes of women over the age of 60 towards menopausal symptoms. We arrived by mail. But only women who had undergone surgery to force them into the menopause responded to the survey. The healthy ones did not want to take the time to fill it in. They deliberately avoided answering. In this case, can we reach the general opinion of women over 60? If we always get answers from those who have undergone surgical menopause, could there be very negative attitudinal statements with the bias of suddenly entering the menopause without being psychologically ready? Do you think they are still used? Survey Methods Interviewer bias occurs when the expectations or opinions of the person conducting the interview affect their objectivity, either negatively or positively, and cloud their judgement of the person being interviewed. Interviewer bias usually relates to aspects of the interviewers and the way in which they ask questions and respond to the answers. For example, we can ask the question in two ways like that What do you think is the most important problem with television programmes? Do you think violence in TV programmes is the most important problem? Do you think the second question contains a referral? Additionally, such bias may stem from perceptions of the interviewer’s identity. The interviewer’s sex, ethnicity, age, attractiveness, social class, level of education, or professional background may affect how participants respond to questions, especially where these characteristics seemingly relate to the interview topic. Survey Methods Selection bias is a systematic error due to differences between those who choose to participate in studies and those who do not. Selection bias distorts data and leads to unreliable research outcomes. For example, suppose you’re researching the career preferences of final-year students. If only female students volunteer to participate in your study, then your data could be affected by volunteer bias (self-selection bias). Another instance of selection bias is when your study applies to people of all income levels but you only have participants from the economically advantaged class → sampling bias In telephone or internet surveys, you can only access those who have a phone or internet. Maybe these people are at a certain socioeconomic level. Low-income families may not have internet. Survey Methods  Survey Methods: Summary ▪ Response rate bias: It can be a threat to all survey methods. ▪ Selection bias: It can be a threat to telephone and internet surveys. ▪ Interviewer bias: It can be a threat to personal and telephone surveys. ▪ Low cost: Internet & Telephone ▪ High confidentiality: Mail & Internet ▪ High time & labour savings: Internet.  Sampling in Survey Research ▪ Careful selection of a survey sample allows researchers to generalize findings from the sample to the population. ▪ The ability to generalize from a sample to the population depends critically on the representativeness of the sample. ▪ A biased sample is one in which the characteristics of the sample are systematically different from the characteristics of the population. Population is the set of all cases of interest. For example, if you are interested in the attitudes of students on your campus toward computer services, your population is all students on your campus We need to develop a specific list of the members of the population to select a subset of that population. This specific list is called a sampling frame. In a survey of students’ attitudes toward computer services, the sampling frame might be a list obtained from the registrar’s office of all currently enrolled students. The subset of the population actually drawn from the sampling frame is called the sample. We might select 100 students from the registrar’s list to serve as the sample for our computer survey Each member of the population is called an element  Sampling in Survey Research Sampling Methods PROBABILITY SAMPLING NONPROBABILITY SAMPLING ▪ Simple Random Sampling ▪ Convenience Sampling ▪ Systematic Sampling ▪ Purposive Sampling ▪ Stratified Sampling ▪ Quota Sampling ▪ Cluster Sampling ▪ Snowball Sampling  Sampling in Survey Research PROBABILITY SAMPLING ▪ The distinguishing characteristic of probability sampling is that the researcher can specify the probability that each element of the population will be included in the sample. ▪ It gives the chance to choose the sample that best represents the population. Sampling in Survey Research Simple Random Sampling Every element has an equal chance of being included in the sample. You list each member of the population and use random numbers to decide which subjects are in the sample  Sampling in Survey Research Cluster Sampling In cluster sampling, the population is divided into clusters which are then chosen at random. Data will be collected from a student dormitory with 10 blocks: 1) Choose 2 randomly out of 10 blocks, 2) Choose 2 floors randomly from these 2 blocks, each of which has 5 floors. 3) Choose 3 rooms randomly from these floors, each of which has 6 rooms. 4) Do the study with the 12 designated rooms. Sampling in Survey Research Stratified Sampling When the population consists of potential sub-populations that will affect the dependent variable, it is the random selection of individuals from these strata Stratified sample seems like cluster sampling, but the strata, or groups, are chosen specifically to represent different characteristics within the population such as ethnicity, age, location or occupation. There are two general ways to determine how many elements should be drawn from each stratum: One way is to draw equal-sized samples from each stratum The second way is to draw elements for the sample on a proportional basis. Consider a population of undergraduate students made up of 30% freshmen, 30% sophomores, 20% juniors, and 20% seniors (class years are the strata). A proportional stratified sample of 200 students drawn from this population would include 60 freshmen, 60 sophomores, 40 juniors, and 40 seniors. In contrast, in disproportional stratified sample, drawing equal-sized samples from each stratum would result in 50 students for each class year. Sampling in Survey Research Systematic Sampling In systematic sampling, sample members from a larger population are selected according to a random starting point but with a fixed, periodic interval This interval, called the sampling interval, is calculated by dividing the population size by the desired sample size. https://www.youtube.com/watch?v=9PaR1TsvnJs  Sampling in Survey Research NONPROBABILITY SAMPLING ▪ Nonprobability sampling does not guarantee that every element in the population has an equal chance of being included in the sample. ▪ It is easy and economical. For example, your population is all students on your campus. If a researcher interviewed the first 30 students who entered the library, (s)he would be using nonprobability sampling. Clearly, not all students would be equally likely to be at the library at that particular time, and some students would have essentially no chance of being included in the sample (e.g., if at work or in class).  Sampling in Survey Research Convenience Sampling = Accidental Sampling This type of sampling is sometimes called accidental sampling. Participants are selected because they are spatially close to and there at the time the researcher collects data. Totally accidental or haphazard. Selecting ready and easily accessible participants: Inclusion the sample members who meet criteria such as easy access, geographical proximity, availability at a given time and willingness to participate. Administering a survey to the students who come to the class Having people who eat at Polin cafe fill out a questionnaire.  Sampling in Survey Research Purposive Sampling In this sampling method, sample is selected based on knowledge about the study and the population. Participants are selected based on the purpose of the sample. The subjects who dont meet the profile are rejected. I may want to study with women who have lost their mother in the last 5 years and are older than 30?  Sampling in Survey Research Quota Sampling In quota sampling, participants or locations are selected nonrandomly according to a fixed quota or percentage of the population based on one or more characteristics. For example: If it is known that 40% of people from Ankara were born in Ankara and 60% are from a To include my students in the study according different city, we can try to select the sample to percentages, whether they were born in according to this percentage. Ankara or not. Since this percentage can change from moment to moment, it cannot be based on probability!  Sampling in Survey Research Snowball Sampling Snowball sampling (or chain-referral sampling) is a sampling method used when characteristics to be possessed by samples are rare and difficult to find. Establish a contact with one or two initial cases from the sampling For example, you might study with frame. This stage is usually the most difficult one. certain ethnic or religious groups, 1) Request the initial cases to identify more cases illegal immigrants, unregistered 2) Ask new cases to identify further cases (and so on) workers etc., it is not possible to 3) Stop when: determine who is included in these a) Your predetermined sample size has been completed, or groups or to estimate the size of the b) There are no further cases left population https://www.youtube.com/watch?v=TtcCvy-CKLc&t=58s  Survey Research Desings 1.Cross-Sectional 2. Longitudinal design 3.Successive independent samples design 4.Cohort sequential samples design  Survey Research Desings 1.Cross-Sectional ▪ One or more samples are drawn from the population(s) simultaneously. ▪ Cross-sectional designs are ideally suited for the descriptive and predictive goals of survey research. ▪ The outcome describes the state of the sample at that time. ✓ e.g. To examine 3-4-5 year old children for language development. ▪ The most important advantage is cost time. ▪ The most important problem is the cohort effect. SAME TIME DIFFERENT SUBJECTS A cohort is a group of people who share a common set of demographic characteristics or experiences, including but not limited to age. Cohort effects describe how studying populations in different “cohorts” — having been born in a different time or region or having different life experiences — can alter the outcomes of studies. For example: In 2019, we looked at the disease rates of 4-5-6 year old groups and observed an incredible increase at age 6. So we can interpret that "diseases increase at the age of 6". Those who were 6 years old in 2019 may have been exposed to high levels of radioactive material in the year they were born and may therefore be more sick. We may miss this cohort effect with a cross- sectional design  Survey Research Desings 2. Longitudinal design ▪ The same respondents are surveyed over time to examine changes in individual respondents. ▪ It is used to see the change over time in the same participants. ✓ e.g. take the same people at age 10, age 20, and age 30 to examine the relationship between intelligence and age. ▪ Two advantages: ▪ The investigator can determine the direction and extent of change for individual respondents. ▪ It is the best survey design for a researcher to assess the effect of some naturally occurring event. DIFFRENT TIME SAME SUBJECTS  Survey Research Desings 2. Longitudinal design ▪ Disadvantages: ▪ This design is incredibly impractical (hugely time-consuming & costly) ▪ Participants may drop out → Attrition is probably the most serious disadvantage of the longitudinal design because as samples decrease over time, they are less likely to represent the original population from which the sample was drawn. ▪ For the sake of consistency, participants might give the same answers, even if they have changed. ▪ Change may occur in the second interview because the first interview sensitized the participants. A period effect is a change which occurs at a particular time, affecting all age groups and cohorts equally. Period effects result from external factors equally affecting all age groups at a particular calendar time. It could arise from various environmental, social and economic factors e.g. war, famine, and economic crisis. For example, we want to observe children when they are 4-5-6 years old and examine whether there are any age-related changes in their aggressive behavior. Measurements were taken in 2018 and 2019 when they were 4 and 5 years old respectively. The last measurement was taken in 2020, with a pandemic in 2020. Aggressive behaviour and tendency to violence increased in children who were constantly at home. According to this longitudinal design, it can be interpreted that behavioural disorders increase at the age of 6! However, the same change was in everyone, not only six-year-old children.  Survey Research Desings 3.Successive independent samples design ▪ Different samples of respondents from the population complete the survey over a while. ▪ It allows researchers to study changes in a population over time. ▪ It does not allow researchers to infer how individual respondents have changed over time. ✓ e.g. Students are asked about their professional goals in 2005, 2010 and 2015. Thus, the changes in the goals of the youth are monitored. ▪ A problem with the successive independent samples design occurs when the samples drawn from the population are not comparable—that is, not equally representative of the population. DIFFERENT TIME DIFFERENT SUBJECTS  Survey Research Desings 4.Cohort sequential samples design ▪ This method consists of limited repeated measurements of independent age cohorts, resulting in temporally overlapping measurements of the various age groups. ▪ It requires less time than the longitudinal design. → shorter follow-up period ▪ By assessing individuals of different ages on multiple occasions, cohort- sequential designs permit researchers to consider cohort and period effects in addition to age changes. Age effects are variations linked to biological and social processes of aging specific to individuals → in research, DIFFRENT TIME any outcome associated with being a certain age! SAME AND DIFFERENT SUBJECTS  Survey Research Desings 4.Cohort sequential samples design ▪ This method consists of limited repeated measurements of independent age cohorts, resulting in temporally overlapping measurements of the various age groups. Longitudinal design Cross-sectional design  Survey Research Desings: Summary ▪ Cross-Sequential: Cohort effect can be confounding. ▪ Longitudinal: Period effect can be confounding. ▪ Successive independent samples: Period effect can be confounding. ▪ Cohort sequential samples: It gives the age effect by considering the cohort and period effects. Research Methods in Psychology-II Özlem Ertan-Kaya, PhD [email protected]