The article that I chose to read on 538 is called “What Do Men Think It Means To Be A Man?”. The article is related to what men think about masculinity in light of the #MeToo movement and societal pressure on gender roles. The data used in the article is actually a survey in which 1600 men took part in to answer questions related to masculinity across multiple dimensions include:
Based on my assessment of the article, I was curious to learn more about how men in the survey felt when it comes to concerns regarding their physical health as that’s an important factor to me.
Before investigating the question of interest to me, I wanted to get
a quick preview of the survey data that FiveThirtyEight used in the
article. In order to obtain this data, I managed to find an associated
.csv file linked on the FiveThirtyEight article page on the
website which redirected me to a GitHub
repository.
# Load the survey data from 538
url <- "https://raw.githubusercontent.com/fivethirtyeight/data/refs/heads/master/masculinity-survey/masculinity-survey.csv"
# Create a DataFrame from the URL
df <- read.csv(url)
head(df)## AMONG.ADULT.MEN X
## 1
## 2 In general, how masculine or "manly" do you feel?
## 3 Very masculine
## 4 Somewhat masculine
## 5 Not very masculine
## 6 Not at all masculine
## Adult.Men Age X.1 X.2 Race X.3 Children X.4
## 1 18 - 34 35 - 64 65 and up White Non-white Has children No children
## 2
## 3 37% 29% 42% 37% 34% 44% 43% 30%
## 4 46% 47% 46% 47% 50% 39% 47% 46%
## 5 11% 13% 9% 13% 11% 11% 6% 17%
## 6 5% 10% 2% 3% 4% 6% 3% 6%
## Sexual.Orientation X.5
## 1 Straight Gay/Bisexual
## 2
## 3 40% 21%
## 4 47% 49%
## 5 10% 19%
## 6 3% 10%Based on the initial preview of the data, I felt that the survey
questions, options, and responses were really hard to decipher based on
how the Data Frame was already presented. In order to make the entire
Data Frame more interpretative, I wanted to create a data structure that
allows me to easily map the various questions asked in the survey to the
options and corresponding responses to those options for each survey
respondent. Since this data structure seems like a reference table, I
figured the most suitable built-in data structure in R to use was the
list. Therefore, I proceeded to create a list where the
questions are the keys and the responses are the values. In order to
make this transformation reusable, I created a R function called
list_transform which allows me to pass in any Data Frame, a
column to use as the key mapper, and the specific columns I want to
leverage for mapping the values to the keys.
list_transform <- function(df, key_col, value_cols) {
# Ensure key_col and value_cols exist in the dataframe
if (!(key_col %in% colnames(df))) {
stop("Invalid key column name provided.")
}
if (!all(value_cols %in% colnames(df))) {
stop("One or more value columns do not exist in the dataframe.")
}
# Identify rows where the key column is not empty or NA
key_indices <- which(df[[key_col]] != "" & !is.na(df[[key_col]]))
# Initialize named list
result_list <- list()
for (i in seq_along(key_indices)) {
start_idx <- key_indices[i]
end_idx <- ifelse(i < length(key_indices), key_indices[i + 1] - 1, nrow(df))
# Extract relevant rows for the value columns
subset_df <- df[start_idx:end_idx, value_cols, drop = FALSE]
# Remove completely empty rows
subset_df <- subset_df[rowSums(is.na(subset_df) | subset_df == "") < length(value_cols), , drop = FALSE]
key <- df[[key_col]][start_idx]
# Append values for the same key
if (key %in% names(result_list)) {
result_list[[key]] <- bind_rows(result_list[[key]], subset_df) }
else {
result_list[[key]] <- subset_df
}
}
return(result_list)
}I then renamed the first two columns to refer to the
Questions and Options associated with each
question before applying the list_transform function on the
Questons column as the key and an array containing
Options, Adult.Men, and Age as
the value columns. I also validated whether the individual questions
were showing up as keys by using the names() function. In
order to avoid typing the question strictly as a string when trying to
refer to a particular set of questions and options, I ended up
leveraging an index on the key_names variable to easily
have a numerical way of getting the exact string I needed when
referencing the list. I also transformed this into a R function for re
usability.
# Rename the first two columns
colnames(df)[1:2] <- c('Questions', 'Options')
# Apply the list_transform to the first two columns
ref_list <- list_transform(df, 'Questions', c('Options', 'Adult.Men', 'Age'))
print(ref_list)## $`In general, how masculine or "manly" do you feel?`
## Options Adult.Men Age
## 3 Very masculine 37% 29%
## 4 Somewhat masculine 46% 47%
## 5 Not very masculine 11% 13%
## 6 Not at all masculine 5% 10%
## 7 No answer 1% 0%
##
## $`How important is it to you that others see you as masculine?`
## Options Adult.Men Age
## 9 Very important 16% 18%
## 10 Somewhat important 37% 38%
## 11 Not too important 28% 18%
## 12 Not at all important 18% 26%
## 13 No answer 0% 0%
##
## $`Where have you gotten your ideas about what it means to be a good man? (Select all that apply.)`
## Options Adult.Men Age
## 15 Father or father figure(s) 64% 58%
## 16 Mother or mother figure(s) 41% 42%
## 17 Friends 35% 42%
## 18 Other family members 35% 37%
## 19 Pop culture 23% 42%
## 20 Other (please specify) 17% 11%
##
## $`Do you think that society puts pressure on men in a way that is unhealthy or bad for them?`
## Options Adult.Men Age
## 22 Yes 60% 70%
## 23 No 39% 29%
## 24 No answer 1% 0%
##
## $`Ask a friend for professional advice`
## Options Adult.Men Age
## 26 Often 15% 26%
## 27 Sometimes 42% 36%
## 28 Rarely 23% 13%
## 29 Never, but open to it 12% 15%
## 30 Never, and not open to it 7% 9%
## 31 No answer 1% 0%
##
## $`Ask a friend for personal advice`
## Options Adult.Men Age
## 33 Often 14% 22%
## 34 Sometimes 41% 39%
## 35 Rarely 27% 17%
## 36 Never, but open to it 10% 11%
## 37 Never, and not open to it 7% 9%
## 38 No answer 1% 2%
##
## $`Express physical affection to male friends, like hugging, rubbing shoulders`
## Options Adult.Men Age
## 40 Often 14% 21%
## 41 Sometimes 28% 25%
## 42 Rarely 25% 19%
## 43 Never, but open to it 10% 14%
## 44 Never, and not open to it 22% 21%
## 45 No answer 1% 0%
##
## $Cry
## Options Adult.Men Age
## 47 Often 7% 13%
## 48 Sometimes 33% 35%
## 49 Rarely 38% 30%
## 50 Never, but open to it 10% 8%
## 51 Never, and not open to it 9% 14%
## 52 No answer 4% 0%
##
## $`Get in a physical fight with another person`
## Options Adult.Men Age
## 54 Often 2% 5%
## 55 Sometimes 5% 10%
## 56 Rarely 23% 22%
## 57 Never, but open to it 19% 19%
## 58 Never, and not open to it 50% 43%
## 59 No answer 1% 2%
##
## $`Have sexual relations with women, including anything from kissing to sex`
## Options Adult.Men Age
## 61 Often 37% 41%
## 62 Sometimes 26% 11%
## 63 Rarely 13% 14%
## 64 Never, but open to it 10% 19%
## 65 Never, and not open to it 11% 16%
## 66 No answer 2% 0%
##
## $`Have sexual relations with men, including anything from kissing to sex`
## Options Adult.Men Age
## 68 Often 5% 10%
## 69 Sometimes 4% 6%
## 70 Rarely 3% 6%
## 71 Never, but open to it 6% 12%
## 72 Never, and not open to it 80% 65%
## 73 No answer 1% 0%
##
## $`Watch sports of any kind`
## Options Adult.Men Age
## 75 Often 38% 29%
## 76 Sometimes 25% 19%
## 77 Rarely 18% 24%
## 78 Never, but open to it 5% 7%
## 79 Never, and not open to it 13% 20%
## 80 No answer 1% 0%
##
## $`Work out`
## Options Adult.Men Age
## 82 Often 21% 30%
## 83 Sometimes 32% 31%
## 84 Rarely 28% 25%
## 85 Never, but open to it 10% 7%
## 86 Never, and not open to it 7% 7%
## 87 No answer 2% 0%
##
## $`See a therapist`
## Options Adult.Men Age
## 89 Often 5% 9%
## 90 Sometimes 9% 10%
## 91 Rarely 15% 11%
## 92 Never, but open to it 37% 35%
## 93 Never, and not open to it 32% 36%
## 94 No answer 2% 0%
##
## $`Feel lonely or isolated`
## Options Adult.Men Age
## 96 Often 17% 24%
## 97 Sometimes 29% 32%
## 98 Rarely 31% 26%
## 99 Never, but open to it 5% 3%
## 100 Never, and not open to it 15% 15%
## 101 No answer 2% 0%
##
## $`Which of the following do you worry about on a daily or near daily basis? (Select all that apply.)`
## Options
## 103 Your weight
## 104 Your finances, including your current or future income, assets, or debt
## 105 Your physical health
## 106 Your physique
## 107 Your ability to provide for your family, current or anticipated
## 108 Your mental health
## 109 Sexual performance or amount of sex
## 110 Your hair or hairline
## 111 Your clothing or style
## 112 Appearance of your genitalia
## 113 Your height
## 114 None of the above
## Adult.Men Age
## 103 54% 45%
## 104 53% 53%
## 105 49% 42%
## 106 33% 40%
## 107 32% 33%
## 108 32% 45%
## 109 23% 24%
## 110 19% 33%
## 111 18% 31%
## 112 13% 17%
## 113 6% 12%
## 114 13% 10%
##
## $`Which of the following categories best describes your employment status?`
## Options Adult.Men Age
## 116 Employed, working full-time 48% 40%
## 117 Employed, working part-time 10% 14%
## 118 Not employed, student 6% 18%
## 119 Not employed-retired 19% 4%
## 120 Not employed, looking for work 9% 19%
## 121 Not employed, NOT looking for work 8% 5%
## 122 No answer 0% 0%
##
## $`AMONG EMPLOYED: In which of the following ways would you say it?s an advantage to be a man at your work right now? (Select all that apply).`
## Options Adult.Men
## 124 Men are taken more seriously 23%
## 125 Men make more money 18%
## 126 Men have more choice 16%
## 127 Men have more promotion/professional development opportunities 14%
## 128 Men generally have more support from their managers 11%
## 129 Men are explicitly praised more often 8%
## 130 Other (please specify) 7%
## 131 None of the above 59%
## Age
## 124
## 125
## 126
## 127
## 128
## 129
## 130
## 131
##
## $`AMONG EMPLOYED: In which of the following ways would you say it?s a disadvantage to be a man at your work right now? (Select all that apply.)`
## Options Adult.Men Age
## 133 Managers want to hire and promote women 18%
## 134 Greater risk of being accused of sexual harassment 42%
## 135 Greater risk of being accused of being sexist or racist 38%
## 136 Other (please specify) 5%
## 137 None of the above 42%
##
## $`AMONG EMPLOYED: Have you seen or heard of a sexual harassment incident at your work? If so, how did you respond? (Select all that apply.)`
## Options Adult.Men Age
## 139 Contacted the HR department 8%
## 140 Did not respond at all 8%
## 141 Reached out to the victim to offer support 7%
## 142 Contacted the manager of the accused person 7%
## 143 Confronted the accused person 5%
## 144 Other (please specify) 3%
## 145 Never witnessed sexual harassment 75%
##
## $`AMONG THOSE WHO DID NOT RESPOND TO HARASSMENT: And which of the following is the main reason you did not respond?`
## Options Adult.Men Age
## 147 You didn't want to get involved
## 148 You weren't sure it was sexual harassment
## 149 You didn't think it was your place
## 150 You weren't sure who to contact
## 151 Other (please specify)
## 152 No answer
##
## $`AMONG EMPLOYED: How much have you heard about the #MeToo movement?`
## Options Adult.Men Age
## 154 A lot 42%
## 155 Some 24%
## 156 Only a little 11%
## 157 Nothing at all 24%
## 158 No answer 0%
##
## $`AMONG EMPLOYED WHO'VE HEARD OF METOO: As a man, would you say you think about your behavior at work differently in the wake of #MeToo?`
## Options Adult.Men Age
## 160 Yes 34%
## 161 No 65%
## 162 No answer 1%
##
## $`Do you typically feel as though you're expected to make the first move in romantic relationships?`
## Options Adult.Men Age
## 164 Yes 61% 62%
## 165 No 38% 38%
## 166 No answer 1% 1%
##
## $`How often do you try to be the one who pays when on a date?`
## Options Adult.Men Age
## 168 Always 49% 36%
## 169 Often 25% 22%
## 170 Sometimes 16% 24%
## 171 Rarely 2% 3%
## 172 Never 7% 12%
## 173 No answer 2% 2%
##
## $`AMONG THOSE WHO TRY TO PAY ON FIRST DATE: Which of the following are reasons why you try to pay on dates? (Select all that apply).`
## Options
## 175 It's the right thing to do
## 176 You asked the person out, so you feel obligated to pay
## 177 You feel good about being the one who pays
## 178 Societal expectations, even though you don?t always think you should have to pay
## 179 You make more money than your date
## 180 You try to pay to see if the other person will offer to share in the cost
## 181 Other (please specify)
## Adult.Men Age
## 175 66%
## 176 49%
## 177 39%
## 178 27%
## 179 11%
## 180 7%
## 181 7%
##
## $`When you want to be physically intimate with someone, how do you gauge their interest? (Select all that apply.)`
## Options Adult.Men
## 183 Every situation is different 59%
## 184 Read their physical body language to see if they are interested 46%
## 185 It isn?t always clear how to gauge someone?s interest 35%
## 186 Ask for a verbal confirmation of consent 31%
## 187 Make a physical move to see how they react 29%
## 188 Other (please specify) 7%
## Age
## 183 54%
## 184 46%
## 185 37%
## 186 43%
## 187 34%
## 188 6%
##
## $`Over the past 12 months, when it comes to sexual boundaries, which of the following things have you done? (Select all that apply.)`
## Options
## 190 Wondered whether you pushed a partner too far in a past sexual encounter.
## 191 Talked with a friend or friends about whether you pushed a partner too far.
## 192 Contacted a past sexual partner to ask whether you went too far in any of you sexual encounters.
## 193 None of the above
## Adult.Men Age
## 190 10% 15%
## 191 5% 9%
## 192 3% 7%
## 193 85% 76%
##
## $`Have you changed your behavior in romantic relationships in the wake of #MeToo movement?`
## Options Adult.Men Age
## 195 Yes 13% 23%
## 196 No 86% 77%
## 197 No answer 1% 0%
##
## $`Are you now married, widowed, divorced, separated, or have you never been married?`
## Options Adult.Men Age
## 199 Married 46% 15%
## 200 Widowed 3% 0%
## 201 Divorced 13% 7%
## 202 Separated 2% 0%
## 203 Never married 36% 77%
## 204 No answer 0% 0%
##
## $`Do you have any children? (Select all that apply.)`
## Options Adult.Men Age
## 206 Yes, one or more children under 18 21% 21%
## 207 Yes, one or more children 18 or older 37% 6%
## 208 No children 46% 74%
##
## $`Would you describe your sexual orientation as:`
## Options Adult.Men Age
## 210 Straight 84% 73%
## 211 Gay 8% 12%
## 212 Bisexual 3% 6%
## 213 Other 3% 5%
## 214 No answer 2% 4%
##
## $`Are you:`
## Options Adult.Men Age
## 216 White 67% 52%
## 217 Black 11% 19%
## 218 Hispanic 14% 16%
## 219 Asian 3% 6%
## 220 Other 5% 7%
##
## $`What is the last grade of school you completed?`
## Options Adult.Men Age
## 222 Did not complete high school 4% 3%
## 223 High school or G.E.D. 36% 46%
## 224 Associate's degree 9% 5%
## 225 Some college 21% 21%
## 226 College graduate 27% 25%
## 227 Post graduate degree 2% 1%
##
## $age3
## Options Adult.Men Age
## 229 18 - 34 29% 100%
## 230 35 - 64 52% 0%
## 231 65 and up 19% 0%# Key Names
key_names <- names(ref_list)
print(key_names)## [1] "In general, how masculine or \"manly\" do you feel?"
## [2] "How important is it to you that others see you as masculine?"
## [3] "Where have you gotten your ideas about what it means to be a good man? (Select all that apply.)"
## [4] "Do you think that society puts pressure on men in a way that is unhealthy or bad for them?"
## [5] "Ask a friend for professional advice"
## [6] "Ask a friend for personal advice"
## [7] "Express physical affection to male friends, like hugging, rubbing shoulders"
## [8] "Cry"
## [9] "Get in a physical fight with another person"
## [10] "Have sexual relations with women, including anything from kissing to sex"
## [11] "Have sexual relations with men, including anything from kissing to sex"
## [12] "Watch sports of any kind"
## [13] "Work out"
## [14] "See a therapist"
## [15] "Feel lonely or isolated"
## [16] "Which of the following do you worry about on a daily or near daily basis? (Select all that apply.)"
## [17] "Which of the following categories best describes your employment status?"
## [18] "AMONG EMPLOYED: In which of the following ways would you say it?s an advantage to be a man at your work right now? (Select all that apply)."
## [19] "AMONG EMPLOYED: In which of the following ways would you say it?s a disadvantage to be a man at your work right now? (Select all that apply.)"
## [20] "AMONG EMPLOYED: Have you seen or heard of a sexual harassment incident at your work? If so, how did you respond? (Select all that apply.)"
## [21] "AMONG THOSE WHO DID NOT RESPOND TO HARASSMENT: And which of the following is the main reason you did not respond?"
## [22] "AMONG EMPLOYED: How much have you heard about the #MeToo movement?"
## [23] "AMONG EMPLOYED WHO'VE HEARD OF METOO: As a man, would you say you think about your behavior at work differently in the wake of #MeToo?"
## [24] "Do you typically feel as though you're expected to make the first move in romantic relationships?"
## [25] "How often do you try to be the one who pays when on a date?"
## [26] "AMONG THOSE WHO TRY TO PAY ON FIRST DATE: Which of the following are reasons why you try to pay on dates? (Select all that apply)."
## [27] "When you want to be physically intimate with someone, how do you gauge their interest? (Select all that apply.)"
## [28] "Over the past 12 months, when it comes to sexual boundaries, which of the following things have you done? (Select all that apply.)"
## [29] "Have you changed your behavior in romantic relationships in the wake of #MeToo movement?"
## [30] "Are you now married, widowed, divorced, separated, or have you never been married?"
## [31] "Do you have any children? (Select all that apply.)"
## [32] "Would you describe your sexual orientation as:"
## [33] "Are you:"
## [34] "What is the last grade of school you completed?"
## [35] "age3"# Get the index of choice
col_to_idx <- key_names[34]
print(col_to_idx)## [1] "What is the last grade of school you completed?"# Function of above workflow
key_value_view <- function(key_names, idx, kv_map) {
# Check if idx is valid
if (idx >= length(key_names) || idx < 1) {
stop("Index is out of range for key_names.")
}
# Get the exact key
key_to_select <- key_names[idx]
# Check if the key exists in kv_map
if (!key_to_select %in% names(kv_map)) {
stop("Key not found in kv_map.")
}
return(kv_map[[key_to_select]]) # Use double brackets for list extraction
}In order to flexibly get specific question / option pairs, I wanted
to create a mechanism that allows me to type in a keyword of interest to
find the respondent data associated with my question much faster rather
than sifting through each question / options pair one-by-one. To achieve
this, I needed to come up with a way to perform sub-string search within
R and so I ended up using the grepl() function which can
perform sub-string search and then I built a function in R that sifts
through the key names and finds which key name or hits of key names
contain the keyword that I was interested to find. I then ended up using
this mechanism to search for physical health in order to
get the answer to my original question.
keyword_search <- function(key_names, keyword) {
# Find indices where the substring is found
indices <- which(grepl(keyword, key_names, ignore.case=TRUE))
# Return the indices (0 if no match)
if (length(indices) == 0) {
return(0)
} else {
return(c(key_names[indices], indices))
}
}
# Function Call
result <- key_value_view(key_names, 34, ref_list)
print(result)## Options Adult.Men Age
## 222 Did not complete high school 4% 3%
## 223 High school or G.E.D. 36% 46%
## 224 Associate's degree 9% 5%
## 225 Some college 21% 21%
## 226 College graduate 27% 25%
## 227 Post graduate degree 2% 1%# Key Names Associated with the Options
option_key_names <- df$Options
# Keyword Search
example_search <- keyword_search(option_key_names, 'physical health')
print(example_search)## [1] "Your physical health" "105"# Row in the DataFrame where the physical health option
df_row = df[example_search[2],]
print(df_row$Adult.Men)## [1] "49%"Based on the analysis above, I now know that 49% of
the Adult Men respondents of the survey expressed concern or worry over
their physical health. I wish there was more qualitative information to
break down the reasons for concern or worry over physical health since
the emphasis on the survey was not solely based on masculinity, but also
based on its association post #MeToo movement and societal pressures. I
hypothesize that there is increasing societal pressure for men to look
and feel a certain way as physical health may have a lot to do with
appearance, but then I also saw that appearance was a
separate category which was further separated by various facets such as
hairline and genitalia so clearly the physical health may have been too
broad to make a proper assessment.
One recommendation I have for the survey creators is to consider breaking down physical health into components similar to some of the other categories / subcategories that exist within Concerns and Worries so it would be possible to better understand what is the specific aspect (s) of physical health that generates concern or worry for men and what that essentially means for their masculinity.